Temperature Effect On Prestressed Integral Bridge Beam

PSZ 19:16 (Pind.
1/07)
UNIVERSITI TEKNOLOGI MALAYSIA

DECLARATION OF THESIS / UNDERGRADUATE PROJECT PAPER AND COPYRIGHT
Authors full name :
MUHAMMAD LUTFI BIN OTHMAN
Date of birth
1 JULY 1985
Title
EFFECT OF TEMPERATURE ON PRESTRESSED INTEGRAL

BRIDGE BEAM
Academic Session :
2008/2009
I declare that this thesis is classified as:
CONFIDENTIAL
(Contains confidential information under the Official Secret

Act 1972)*
RESTRICTED
(Contains restricted information as specified by the

organization where research was done)*
OPEN ACCESS
I agree that my thesis to be published as online open access

(full text)
I acknowledged that Universiti Teknologi Malaysia reserves the right as follows:

1. The thesis is the property of Universiti Teknologi Malaysia.
2. The Library of Universiti Teknologi Malaysia has the right to make copies for the purpose
of research only.
3. The Library has the right to make copies of the thesis for academic exchange.
Certified by :
(WRITERS SIGNATURE)
(SIGNATURE OF SUPERVISOR)
(850701-01-6517)
DR. REDZUAN BIN ABDULLAH
Date : 23 NOVEMBER 2009
NOTES :
Date : 23 NOVEMBER 2009
If the thesis is CONFIDENTAL or RESTRICTED, please attach with the letter from
the organization with period and reasons for confidentiality or restriction.
I hereby declare that I have read this project report

and in my opinion this thesis is sufficient terms
of scope and quality for the award of the degree
of Master Engineering
Signature
Name of Supervisor: DR. REDZUAN BIN ABDULLAH

Date
THE EFFECT OF TEMPERATURE ON PRESTRESSED INTEGRAL

BRIDGE BEAM
MUHAMMAD LUTFI BIN OTHMAN
A report submitted in partial fulfillment of the

requirements for the award of the degree of
Master of Engineering (Civil Structure)
Faculty of Civil Engineering

Universiti Teknologi Malaysia
NOVEMBER 2009
ii
I declare that this report entitled The Effect of Temperature on Prestressed Integral
Bridge Beam is a result of my own research except as cited in the references. The
research has not been accepted for any degree and is not currently concurrently
submitted in candidature of any other degree
Signature
Name
: MUHAMMAD LUTFI BIN OTHMAN
Date
iii
Karya ini adalah dedikasi teristimewa buat emak yang

dikasihi, Hamidah Bt Ahmad, abah yang disayangi, Othman Bin Mustaffa serta adik
tercinta, Fatimah Azzahrah Bt Othman yang tidak pernah jemu membekalkan nasihat,
kekuatan dan semangat untukku menghadapi liku-liku hidup seorang mahasiswa.
Tidak lupa juga buat Maktok, Allahyarham Abahtok, Tok Mah,
Tok Mat, sanak saudara serta semua teman-teman
seperjuanganku di Universiti Teknologi Malaysia
iv
ACKNOWLEDGEMENT
Praise be to Allah S.W.T, after months of hard work and brainstorming, this
masters project entitled The Effect of Temperature on Prestressed Integral Bridge
Beam is finally completed. Thanks to Allah S.W.T as to his guidance and mercy, this
thesis can at last be finished within the allocated time.
In this opportunity, I would like to express my gratitude towards my supervisor

for this project, Dr. Redzuan Bin Abdullah for his advice and kindness in guiding me
and my partner throughout the semester. Only Allah S.W.T can repay your kindness. I
would also like to give my sincerest thanks to Ir. Mohamad Salleh Bin Yassin for his
brilliant ideas, supportive critics, and also for being a huge helping hand in time of
needs. On top of that, I would like to give my special thanks to my partner, Mohd
Fairuz Omar for all the cooperation, help, and unwavering commitment throughout the
development of this study.
Last but not least, thanks to my mother, father, sister, all my family members,
all my friends and all the individuals for the moral support given.
ABSTRACT
Most bridges in Malaysia are pre-stressed concrete type and constructed as

simply supported beam. Such structural form has maintenance problems due to the
existence of joints and bearings which are easily deteriorated. Therefore, the
elimination or the minimizing of bearings and joints is very important in order to
increase the durability and life span of bridge structure. The Malaysian authority
(Jabatan Kerja Raya) has now regulated that bridges with span length less than 60m
should be designed and constructed as integral structure. The integral bridge
however, is highly indeterminate and the behavior is affected by the change of
temperature. The differential thermal effects can cause transverse bending at the
middle pear of integral bridge which can result in the uplift. On the other hand, the
continuous construction at the ends may result in crack. This research is aimed to
develop a finite element model for integral bridge using LUSAS software. After the
model is developed and verified, it is used to study the effect of temperature on
prestress force in integral bridge beam. The longer the span of the beam, the more the
temperature will affect the prestress change. Even in Malaysian condition where the
temperature only ranges from 22 C
to 35 C, it is important to consider temperature
effects in designing integral bridge.
vi
ABSTRAK
Kebanyakan jambatan di Malaysia ialah jenis konkrit pra-tegasan dan dibina

sebagai rasuk sokong mudah. Struktur jenis ini kebiasaannya terdedah kepada
masalah penyelenggaraan disebabkan kewujudan sambungan dan bering yang mudah
memburuk. Oleh itu, adalah sangat penting untuk tidak menggunakan sambungan
dan bering, atau setidak-tidaknya meminimakan penggunaannya agar kebolehkerjaan
dan jangka hayat struktur jambatan boleh ditingkatkan. Pihak berkuasa tempatan
(Jabatan Kerja Raya) kini telahpun mensyaratkan supaya jambatan dengan panjang
kurang daripada 60m direkabentuk dan dibina sebagai struktur integral.
Walaubagaimanapun, jambatan integral mempunyai ketidaktentuan yang tinggi
selain dipengaruhi oleh kesan suhu. Kesan suhu yang berbeza boleh menyebabkan
lenturan ke atas pada bahagian tengah jambatan. Disebabkan itu, sambungan integral
pada hujung rasuk mempunyai kemungkinan untuk retak. Penyelidikan ini
mensasarkan pembangunan model unsur terhingga untuk jambatan integral dengan
menggunakan perisian LUSAS. Selepas model tersebut berjaya dibangunkan dan
dibuktikan, model tersebut kemudiannya digunakan untuk mengkaji kesan suhu
terhadap daya pra-tegas pada rasuk jambatan integral. Semakin panjang unjuran
rasuk, semakin besar kesan suhu terhadap perubahan daya pra-tegas. Walaupun suhu
hanya berubah antara 22 C hingga 35 C di Malaysia, kesan suhu masih sangat
penting untuk diambilkira dalam merekabentuk jambatan integral.
vii
TABLE OF CONTENT
CHAP
CONTENT
PAGE
TITLE
DECLARATION
ii
DEDICATION
iii
ACKNOWLEDGEMENT
iv
ABSTRACT
ABSTRAK
vi
TABLE OF CONTENT
vii
LIST OF TABLES
xi
LIST OF FIGURES
xiv
LIST OF NOTATIONS
xviii
LIST OF APPENDICES
xix
INTRODUCTION
1.1
Background Study
1.2
Problem Statement
1.3
Research Objectives
1.4
Research Questions
1.5
Scope of The Research
viii
CHAP
CONTENT
BACKGROUND STUDY OF BRIDGE ANALYSIS
PAGE
AND INTEGRAL BRIDGE
2.1
Bridge
2.2
Integral Bridges
2.2.1
Integral Bridge Issues
2.2.2
Integral Bridge Limitations
2.2.3
Integral Bridge Construction
2.3
Bridge Deck Analysis
12
2.3.1
Simply Supported Beam/Slab
13
2.3.2
Series of Simply Supported Beams/Slabs
14
2.3.3
Continuous Beam/Slab With Full Propping

During Construction
16
2.3.4
Partially Continuous Beam/Slab
16
2.3.5
Frame/Box Culvert (Integral Bridge)
19
2.4
Articulation
23
2.5
Bridge Loading
26
2.5.1
Dead and Superimposed Dead Loading
28
2.5.2
Imposed Traffic Loading
29
2.5.3
Imposed Loading Due to Road Traffic
29
2.5.4
Thermal Loading
30
ix
CHAP
CONTENT
PAGE
RESEARCH METHODOLOGY
34
3.1
Introduction
33
3.2
Problem Identification
36
3.3
Data Collection
36
3.3.1
Previous Research
36
3.3.2
Adaptation of Real Integral Bridge Design
38
3.3.2.1
Model Layout
38
3.3.2.2
Support Connection Details
39
3.4
LUSAS Structural Modelling
40
3.4.1
Assumptions
41
3.4.2
Geometry Definition
42
3.4.3
Attributes Definition
42
3.4.3.1
Meshing Attribute
43
3.4.3.2
Geometric Attribute
45
3.4.3.3
Material Attribute
49
3.4.3.4
Support Attribute
50
3.4.3.4.1
53
Example Calculation
For Spring Stiffness
3.4.3.5
3.4.4
3.5
Loading Attribute
Prestress Definition to BS5400
Prestress Change Determinationby Trial and Error
54
56
57
Method
3.6
Comparison of LUSAS and Hand Calculation Trial
59
and Error Method

3.6.1
Example Calculation for Hand Calculation
60
Method
3.7 Analysis and Result Interpretation
64
CHAP
CONTENT
RESULT AND DISCUSSION
65
4.1
Introduction
65
4.2
Analysis Outline
65
4.3
Comparison of Deflection
65
4.3.1
Case 1: Temperature Gradient (TT>TB)
65
4.3.2
Case 2: Temperature Gradient (TT<TB)
68
4.3.3
Case 3: Uniform Temperature Increment
69
4.3.4
Case 4: Uniform Temperature Decrement
71
4.4
PAGE
Prestress Change in Prestressed Integral Bridge Beam
73
4.4.1
Case 1: Temperature Gradient (TT>TB)
74
4.4.2
Case 2: Temperature Gradient (TT<TB)
75
4.4.3
Case 3: Uniform Temperature Increment
77
4.4.4
Case 4: Uniform Temperature Decrement
79
4.5 Discussion
82
CONCLUSION AND RECOMMENDATION
83
5.1
Conclusion
83
5.2
Recommendation
85
REFERENCES
86
APPENDICES
87
xi
LIST OF TABLES
TABLE NO.
3.1
TITLE
Safety factor calculated according to each bridge
PAGE
39
spans Z1 and Z2
3.2
General details for beam span 20m, 30m and 40m
39
3.3
Section properties for 20m, 30m and 40m integral
46
bridge span
3.4
Material properties
49
3.5
Values of spring stiffness
54
3.6
HA loading details
54
3.7
Trial and error method
59
3.8
Comparison of the reduction percentage between
64
LUSAS and hand calculation

4.1.1
Mid-span deflection comparison for 20m span
66
integral bridge and simply supported bridge due to

temperature gradient (TT>TB)
4.1.2

67
xii
TABLE NO
4.1.3
TITLE
PAGE
67

4.2.1
68

temperature gradient (TT<TB)
4.2.2
68

4.2.3
68

4.3.1
70

uniform temperature increment
4.3.2
70

4.3.3
70

4.4.1
72

uniform temperature decrement
4.4.2
72

4.4.3

72
xiii
TABLE NO
TITLE
PAGE
4.5.1
Prestress decrement percentage due to temperature
74
gradient (TT>TB) for 20m span integral bridge

4.5.2
74

4.5.3
74

4.6.1
76
gradient (TT<TB) for 20m span integral bridge

4.6.2
76

4.6.3
76

4.7.1
Prestress decrement percentage due to uniform
78
temperature increment for 20m span integral

bridge
4.7.2
78

bridge
4.7.3
78

bridge
4.8.1
80
temperature decrement for 20m span integral

bridge
4.8.2
80

bridge
4.8.3

bridge
80
xiv
LIST OF FIGURES
FIGURE NO.
TITLE
PAGE
2.1
Typical design of integral bridge
2.2
(a) Precast beams made integral over the interior
10
support
(b) deck continuous over interior support and
10
integral with abutments
2.3
(c) deck integral with abutments and pier
10
(a) geometry of integral bridge
11
(b) deformed shape if bases are restrained against
11
sliding
(c) bending moment diagram if bases are
11
restrained against sliding

(d) deformed shape if bases are partially restrained
11
against sliding
2.4
Portion of bridge illustrating bridge engineering
13
terms
2.5
Simply supported beam or slab
14
2.6
Series of simply supported beam/slabs
14
2.7
Continuous beam or slab
15
2.8
Bending moment diagrams due to uniform loading
15
of intensity
xv
FIGURE NO.
2.9
TITLE
(a) Elevation view of partially continuous bridge
PAGE
16
with full-depth diaphragm at intermediate supports

(b) Plan view from below of partially continuous
17
bridge with full-depth diaphragm at intermediate

supports
2.10
Partially continuous bridge with continuity
17
provided only by the slab at intermediate supports

2.11
Joint detail at intermediate support of partially-
18
continuous bridge of the type illustrated in figure

2.10
2.12
(a) Bending moment due to selfweight
18
(b) Bending moment due to loading applied after

bridge has been made continuous
18
2.13
Box culvert
19
2.14
Three-span frame
19
2.15
Typical distributions of bending moment
20
2.16
Effect of thermal contraction of deck in frame
21
bridge
2.17
Sliding support and run-on slab in frame bridge
21
2.18
Composite precast and in-situ concrete frame
22
bridge
2.19
Plan views showing articulation of typical bridges
24
2.20
Uplift of bearings due to traffic loading
25
2.21
Uplift of bearing due to transverse loading caused
25
by differential thermal effects

2.22
(a) Beam on sliding bearing
31
(b) Beam fixed at both ends
31
3.1
Methodology flow chart
35
3.2
Overall elevation of the Charles D. Newhouse
37
research test setup

3.3
Diaphragm details
37
xvi
FIGURE NO.
TITLE
PAGE
3.4
Longitudinal integral beam
38
3.5
Integral beam-to abutment connection
39
3.6
Integral beam-to abutment connection detail
40
3.7
Integral bridge beam line
42
3.8
Tendon profile line
42
3.9
Line mesh assignment interface
44
3.10
Active mesh applied to beam and abutments
44
3.11
Arbitrary Section Property Calculator Interface
45
3.12
20m integral bridge span cross-section
46
3.13
47
3.14
48
3.15
Material assignment interface
50
3.16
Structural support setting for roller
51
3.17
Structural support setting for spring stiffness
52
3.18
Visual of roller support and spring stiffness
52
support in LUSAS
3.19
Four conditions of temperature effects
55
3.20
Single tendon prestress assignment interface
56
according to BS5400
3.21
Visual of assigned prestress at tendon
57
3.22
Estimation of the effects of unequal extreme fibre
63
temperatures by the flexibility method

4.1
Temperature gradient (TT>TB) case for deflection
66
comparison
4.2
Graph of mid span deflection versus temperature
67
gradient (TT>TB) for IB and SSB

4.3
Temperature gradient (TT<TB) case for deflection
68
comparison
4.4
Graph of mid span deflection versus temperature

gradient (TT<TB) for IB and SSB
69
xvii
FIGURE NO.
TITLE
PAGE
4.5
Uniform temperature increment case for deflection
69
comparison
4.6
Graph of mid span deflection versus uniform
71
temperature increment for IB and SSB

4.7
Uniform temperature decrement case for
71
deflection comparison
4.8
Graph of mid span deflection versus uniform
73
temperature decrement for IB and SSB

4.9
Temperature gradient (TT>TB) case for prestress
74
change analysis
4.10
Graph of prestress decrement percentage versus
75

4.11
Temperature gradient (TT<TB) case for prestress
75
change analysis
4.12
Graph of prestress increment percentage versus
77

4.13
Uniform temperature increment case for prestress
77
change analysis
4.14
79

4.15
Uniform temperature decrement case for prestress
79
change analysis
4.16

81
xviii
LIST OF NOTATIONS
LUSAS
London University Stress Analysis System

(engineering software)
Dry density of the backfill
Coefficient of thermal expansion
Gs
Specific gravity
Increased temperature
Horizontal stress
Section depth
emax
Eccentricity at mid-span
emin
Eccentricity at support
Area
Moment of Inertia
Moment
TT
Temperature at extreme top of fibre
TB
Temperature at extreme bottom of fibre
Ec
Modulus of Elasticity
fcu
Characteristic strength
Es
Modulus of Elasticity
Concrete creep coefficient
Curvature
Coef.
Coefficient
Temp.
Temperature
IB
Integral bridge
SSB
Simply supported bridge
xix
LIST OF APPENDICES
APPENDIC NO.
TITLE
Design Example of Beam Section and Prestressing

Force by Microsoft Excel
PAGE
87
CHAPTER I
INTRODUCTION
1.1
Background Study
One of the most important structures is bridge. In Malaysia, most of the existing
bridges were design as simple spans. In simple span construction, joints and bearings are parts
of the bridge structure. It is indeed easier for the engineers to design and easier for the
contractors to build simple span bridge but on the other hand, because of the joints between
the spans of the bridge, it will not be able to provide a smooth riding surface to the public and
furthermore a leaking joint will most certainly cause corrosion. The maintenance is also costly
as the bearing needed to be replaced after every few years.
Today, integral bridges have been constructed all over the world instead of the
conventional simple spans bridges. The advantages of integral bridge have been realized as
early as the 60s. The use of integral deck eliminates the need for deck expansion joints and
bearings. More significantly, maintenance costs are also reduced since deck joints, which
allow water to leak onto substructures elements and accelerate deterioration, are totally
eliminated. In addition, future widening or bridge replacement becomes easier, since the
simple design of the integral abutment lends itself to simple structural modification.
2
1.2
Problem Statement
In recent years, it has been established that a significant portion of the world's bridges
are not performing as they should. In some cases, bridges are carrying significantly more
traffic load than originally intended. However, in many others, the problem is one of
durability. This is often associated with joints that are leaking or with details that have
resulted in chloride-contaminated water dripping onto substructures. Problems have also been
reported with post-tensioned concrete bridges in which inadequate grouting of the ducts has
lead to corrosion of the tendons.
The new awareness of the need to design durable bridges has led to dramatic changes
of attitude towards bridge design. There is now a significant move away from bridges that are
easy to design towards bridges that will require little maintenance. The bridges that were easy
to design were usually determinate, e.g. simply supported spans and cantilevers. However,
such structural forms have many joints which are prone to leakage and also have many
bearings which require replacement many times over the lifetime of the bridge.
The move now is towards bridges which are highly indeterminate and which have few
joints or bearings. The structural forms of bridges are closely interlinked with the methods of
construction. The methods of construction in turn are often dictated by the particular
conditions on site. For example, when a bridge is to be located over an inaccessible place,
such as a railway yard or a deep valley, the construction must be carried out without support
from below. This immediately limits the structural forms to those that can be constructed in
this way. The method of construction also influences the distributions of moment and force in
a bridge. For example, in some bridges, steel beams carry the self weight of the deck while
composite steel and in-situ concrete carry the imposed traffic loading.
Integral bridge is advantageous in term of maintenance and long term planning if

compared to the conventional bridge. This type of bridge can also be seen as the future bridge
as it is stiffer and has been observed that the deflection and moments can be greatly reduced
as in case of integral bridge. The elimination or minimizing of bearings and joints is important
as they are fragile elements and represent the weakest links in bridge structures. Joints are
expensive to buy, install, maintain and repair. Sometimes repair costs can run as high as
3
replacement costs. Successive paving will ultimately require that joints be replaced or raised.
Even waterproof joints will leak over time, allowing water salt-laden or otherwise, to pour
through the joints accelerating corrosion damage to girder ends bearings and supporting
reinforced concrete substructures. Accumulated dirt, rocks and trash fill Elastomeric glands
leading to failure.
Bearings are also expensive to buy and install and more costly to replace. Over time,
steel bearings may tip over and seize up due to loss of lubrication or buildup of corrosion.
Elastomeric bearings can split and rupture due to unanticipated movements or ratchet out of
position. Teflon sliding surfaces are fragile for bridge applications and can fail prematurely
due to excessive wear from dirt and other contaminants, or due to poor fabrication and
construction tolerances. Pot bearings also suffer frequent damage due to poor fabrication and
construction techniques.
Integral bridges are characterized by monolithic connection between the deck and the
substructure. Such bridges are the answer for small and medium length bridges where
bearings and joints are either eliminated or reduced to minimum. The integral bridge concept
is an excellent option to incorporating reduced inspection and maintenance features in the
bridge structures. However, it is more complicated to design and the secondary restraint
moments can develop at the connection due to creep, shrinkage, and thermal effects.
In Malaysia, this type of bridge is still not widely used because of its complexity and
the lack of knowledge and experience within Malaysian construction industry.
The purpose of experimental study presented in this paper was to compare the restraint
moments that developed during the early ages of continuity to the predicted restraint moments
using finite element program, LUSAS. It is important to be able to accurately predict the
restraint moment because:
i.
Underprediction leads to unconservative designs and the potential for damage

to cracking at the continuity connection
ii.
Overprediction may force the designer to use simple span design instead of
continuous design
4
1.3
Research Objectives
The objectives of this research are:
1. To study the effect of temperature gradient on the prestressing force in prestressed

integral bridge beam
2. To study the effect of uniform temperature change on the prestressing force in
prestressed integral bridge beam
3. To determine whether temperature effects can be neglected in integral bridge design
considering Malaysian condition
1.4
Research Questions
By the end of this research, it is aimed that the following questions will be answered
1. How integral bridge response to temperature effect?

2. How does the prestress force reacts to temperature loadings?
3. Is it true that integral bridge beam is better than simply supported beam?
4. Can the temperature effects be neglected at certain span of the integral bridge?
1.5
Scope of The Research
In order to finish this research within the limited time, the following scopes are being
considered:
a.
The simulation of the integral bridge will be developed using LUSAS program
and will be verified by consulting superior LUSAS users
5
b.
The temperature gradient is fixed between 0 C to 40 C and the span length

between 20m to 40m
c.
Typical design of integral bridge consists of the beams, piers and abutments is
used
d.
The temperature effect studied only consider Malaysian condition
CHAPTER II
BACKGROUND STUDY OF BRIDGE ANALYSIS AND INTEGRAL

BRIDGE
2.1
Bridge
According to Heinz Kurth in his book, Bridge, a bridge is a permanent raised

structure which allows people or vehicles to cross obstacles such as a river without blocking
the way of traffic passing underneath. In other words, a bridge is a structure built to span a
gorge, valley, road, railroad track, river, body of water, or any other physical obstacle.
Designs of bridges vary depending on the function of the bridge and the nature of the terrain
where the bridge is to be constructed. There are six main types of bridges:
i. Beam bridges
ii. Cantilever bridges
iii. Arch bridges
iv. Suspension bridges
v. Cable-stayed bridges
vi. Truss bridges.
7
2.2
Integral Bridges
According to The State of the Art of Precast/Prestressed Integral Bridges by the

Precast/prestressed Concrete Institute, Chicago, Integral Bridge is a bridge which
accommodate superstructure movements without conventional expansion joints.
Integral bridges consist of superstructures, abutments (also called end bents),

intermediate piers. and foundations. The design of integral bridges is generally similar to that
of conventional bridge design. Special analysis and design considerations required for integral
bridges are primarily associated with the need to accommodate volumetric changes in the
structure, such as thermal movements. Figure below shows the typical design of integral
bridge.
Figure 2.1 : Typical design of integral bridge (Babu P.V.M and Bhandari N.M., 2006)
Integral bridges are the bridges where the superstructure and substructures are
continuous or integral with each other. While the concept is well established, many bridges
built in the 1960s and 1970s were articulated with expansion joints and bearings to separate
the superstructure from the substructure and the surrounding soil. In the 1980s and 1990s,
many of these required rehabilitation due to serviceability problems associated with the joints.
(OBrien et al., 1999)
As a result, integral construction has recently received a great deal of attention and this
form is likely to become much more widespread in the future. In the UK in particular,
designers are now required to consider the use of the integral form for most shorter bridges
(up to 60 m span and 30 skew).
8
2.2.1 Integral Bridge Issues
One of the aims of a designer of concrete bridges should be to reduce maintenance by

minimising the number and sophistication of mechanical engineering devices required for its
operation. Thus, wherever possible, piers should be built into the deck. Where this is not
possible, the next best option is to adopt concrete hinges, then rubber bearings, then fixed
mechanical bearings, and last of all, sliding bearings.
However, the greatest single item of maintenance expenditure on highway bridges is

the expansion joint. Not only do such joints need regular repair and renewal, but they also
allow salt-laden water to attack and corrode the substructure. Most attempts at creating
waterproof joints fail after a number of years of service. Expansion joints also need to be
inspected from beneath, which greatly increases the cost and complexity of abutments.
Integral bridges take the philosophy of mechanical simplicity to its logical conclusion
by either pinning the deck to the abutment or building it in. This eliminates the expansion
joint and greatly simplifies the abutment structure that becomes more like a pier. The
consequence of this type of design is that the abutment either rocks or slides back and forth as
the bridge expands and contracts, causing settlement of the backfill behind the abutment, and
disrupting the road surface. This is overcome either by regular maintenance of the road
surface, or by bridging the disrupted area with a short transition slab. The slab is attached to
the abutment, and so follows its movement, sliding on the substrate. Consequently, a flexible
mastic type joint is required in the blacktop at the end of the transition slab. Some authorities
adopt transition slabs, while others prefer to maintain the road immediately behind the
abutment.
A very economical form of abutment is achieved by separating the functions of soil

retention and support of the deck. The soil is retained by a reinforced earth wall, while the
deck is carried on piles that are allowed to rock within the fill. The piles may be concrete or
steel, and may be encased in pipes to give them the freedom to rock. It needs to be
demonstrated that steel piles would be adequately protected from corrosion; they would
appear to be more at risk than similar piles driven into the embankment. (Benaim et al., 2008)
9
2.2.2 Integral Bridge Limitations
There are limitations on the use of integral bridges. These involve the following factors:
i.
Length of structure. Limitations on length are concerned with passive pressure effects,
stresses in the piles, and the movement capacity of the joints between the approach
slab and the approach pavement. Many state departments of transportation limit the
lengths to 300 ft for steel superstructures and 600 ft for prestressed concrete
superstructures. A few states , like Tennessee, have successfully used longer lengths.
ii.
Structure geometry. Only six states have reported application of integral construction
to curved bridges. Skew angles have generally been below 40 deg. However,
Tennessee has used this method of construction extensively and effectively for curved
bridges as well as bridges with skew angles up to 70 deg.
iii.
Foundations. Integral brtdges require that abutment piles be flexible. Therefore, they
should not be used with pile foundation where rock is closer than 10 ft from the
bottom of the abutment beam unless pre-augered holes for piles are employed . The
New York Department of Transportation specifies a minimum pile penetration of 20 ft
into acceptable soils to ensure adequate flexibility and to provide for scour protection.
The minimum depth is also meant to provide sufficient lateral support for the pile.
particularly when conditions dictate that the top portion of the pile is pre-augered and
back-filled with granular material.
Usually. integral bridges are founded on piles However, there are instances where they
have been supported by spread footings that are founded on rock. They can also be supported
on spread footings on soil if the soil is well compacted and the possibility of settlement of the
foundation is considered in the design. (Precast/Prestressed Concrete Institute, 2001)
2.2.3 Integral Bridge Construction
There are many variations on the basic integral bridge. In figure 2.2 (a), the deck is
composed of separate precast beams in each span. While in the past such a deck might have
10
had a joint over the central support, a more durable form of construction is to make it
continuous over the support using in-situ concrete, as illustrated. In figure 2.2 (b), the deck of
the bridge is continuous over the internal support and integral with the abutments at the ends.
Another type is shown in figure 2.2 (c), this bridge is integral with both the abutments and the
intermediate pier.
Figure 2.2 (a): precast beams made integral over the interior support (OBrien et al., 1999)
Figure 2.2 (b): deck continuous over interior support and integral with abutments
Figure 2.2 (c): deck integral with abutments and pier (OBrien et al., 1999)
While there are considerable durability advantages in removing joints and bearings,
their removal does affect the bridge behaviour. Specifically, expansion and contraction of the
deck is restrained with the result that additional stresses are induced which must be resisted by
the bridge structure. The most obvious cause of expansion or contraction in bridges of all
forms is temperature change but other causes exist, such as shrinkage in concrete bridges. In
prestressed concrete decks, elastic shortening and creep also occur.
A simple integral bridge is illustrated in figure 2.3(a). If the bases of the abutments are
not free to slide, deck contraction induces the deformed shape illustrated in figure 2.3(b) and
the bending moment diagram of figure 2.3(c). Partial sliding restraint at the bases of the
11
abutments results in the deformed shape of figure 2.3(d) and a bending moment diagram
which is similar in shape to that of figure 2.3(c), but of a different magnitude.
Figure 2.3(a): geometry of integral bridge (OBrien et al., 1999)
Figure 2.3 (b): deformed shape if bases are restrained against sliding (OBrien et al., 1999)
Figure 2.3 (c): bending moment diagram if bases are restrained against sliding
Figure 2.3 (d): deformed shape if bases are partially restrained against sliding
Time-dependent contractions in concrete bridge decks induce bending moments in

integral bridges. While the magnitude of creep contraction is time dependent, creep also has
the effect of relieving the induced bending moments overtime. The net effect of this is that
moments induced by creep contraction are small. Shrinkage strain increases with time but the
resulting moments are also reduced by creep.
12
Elastic shortening occurs in post-tensioned prestressed concrete decks during the
application of prestress. If the deck is integral with the supports at the time of stressing,
bending moments are induced. On the other hand, many integral bridges are constructed from
precast pretensioned beams and the bridge is not made integral until after the pretensioning
process is complete. In such cases, no bending moments are induced by the elastic shortening.
Temperature changes are another major source of deck expansion and contraction.
Temperature can be viewed as having a seasonal and hence long-term component as well as a
daily or short-term component. The resistance of an integral bridge to movement of any type
depends largely on the form of construction of the substructures. Three alternative forms are
illustrated in Fig. 4.3. In each case , a run-on slab is shown behind the abutment.
These are commonly placed over the transition zone between the bridge and the
adjacent soil which generally consists of granular backfill material. Figures 4.3(a) and (b)
show two bridges which are integral with high supporting abutments and piled foundations. In
such a case , a reduction in lateral restraint can be achieved by using driven H-section piles
with their weaker axes orientated appropriately. An alternative form of integral construction is
one in which abutments sit on strip foundations like the small bank seat abutment illustrated
in Fig. 4.3 (c). Minimising the sliding resistance at the base of these foundations helps to
reduce the lateral restraint. Care should be taken in the design to ensure that bank seats have
sufficient weight to avoid uplift from applied loads in other spans. (OBrien et al., 1999)
2.3
Bridge Deck Analysis
According to Bridge Deck Analysis by Eugene J. OBrien and Damien L. Keogh, the
main body of the bridge superstructure is known as the deck and can consist of a main part
and cantilevers. The deck spans longitudinally, which is the direction of span, and
transversely, which is perpendicular to it. There may be upstands or downstands at the ends of
the cantilever for aesthetic purposes and to support the parapet which is built to retain the
vehicles on the bridge.
13
It is also said that bridge decks are frequently supported on bearings which transmit the
loads to abutments at the ends or to piers or walls elsewhere. Joints may be present to
facilitate expansion or contraction of the deck at the ends or in the interior. Illustration which
shows the bridge engineering terms is shown in figure 2.4.
Figure 2.4: Portion of bridge illustrating bridge engineering terms

2.3.1 Simply Supported Beam/Slab
The simplest form of bridge is the single-span beam or slab which is simply supported
at its ends. This form is widely used when the bridge crosses a minor road or small river. In
such cases, the span is relatively small and multiple spans are infeasible and/ or unnecessary.
The simply supported bridge is relatively simple to analyse and to construct but is
disadvantaged by having bearings and joints at both ends. The cross-section is often solid
rectangular but can also be voided rectangular, T-section or box-sections. (OBrien et al.,
1999)
14
Figure 2.5: Simply supported beam or slab (OBrien et al., 1999)
2.3.2 Series of Simply Supported Beams/Slabs
When a bridge crossing is too wide for an economical single span, it is possible to
construct what is in effect a series of simply supported bridges, one after the other, as
illustrated in figure 2.6 below:
Figure 2.6 : Series of simply supported beam/slabs (OBrien et al., 1999)
Like single-span bridges, this form is relatively simple to analyse and construct. (It is
particularly favoured on poor soils where differential settlements of supports are anticipated.
It also has the advantage that, if constructed using in-situ concrete, the concrete pours are
moderately sized. In addition, there is less disruption to any traffic that may be below as only
one span needs to be closed at any time . However, there are many joints and bearings with
the result that a series of simply supported beams/ slabs is no longer favoured in practice.
Continuous beams/ slabs, as illustrated in figure 2.7, have significantly fewer joints and
bearings.
15
Figure 2.7: Continuous beam or slab (OBrien et al., 1999)
A further disadvantage of simply supported beam/ slabs in comparison to continuous

ones is that the maximum bending moment in the former is significantly greater than that in
the latter. For example, the bending moment diagrams due to a uniformly distributed loading
of intensity w (kN/m) are illustrated in figure 2.8. It can be seen that the maximum moment in
the simply supported case is significantly greater (about 25%) than that in the continuous
case. The implication of this is that the bridge deck needs to be correspondingly deeper.
(a)
(b)
Figure 2.8: Bending moment diagrams due to uniform loading of intensity, w: (a) 3
simply supported spans of length, l; (b) One 3-span continuous beam with span lengths, l
16
2.3.3 Continuous Beam/Slab with Full Propping During Construction
Continuous beam/slab construction has significant advantages over simply supported

spans in that there are fewer joints and bearings and the applied bending moments are less.
For bridges of moderate total length, the concrete can be poured in-situ in one pour. This
completely removes the need for any joints. However, as the total bridge length becomes
large, the amount of concrete that needs to be cast in one pour can become excessive. This
tends to increase cost as the construction becomes more of a batch process than a continuous
one. (OBrien et al., 1999)
2.3.4 Partially Continuous Beam/Slab
When support from below during construction is expensive or infeasible, it is possible

to use precast concrete or steel beams to construct a partially continuous bridge. Precast
concrete or steel beams are placed initially in a series of simply supported spans. In-situ
concrete is then used to make the finished bridge continuous over intermediate joints. Two
forms of partially continuous bridge are possible. In the form illustrated in figure 2.9, the insitu concrete is cast to the full depth of the bridge overall supports to form what is known as a
diaphragm beam. (OBrien et al., 1999)
Figure 2.9 (a): Elevation view of partially continuous bridge with full-depth diaphragm
at intermediate supports (OBrien et al., 1999)
17
Figure 2.9 (b): Plan view from below of partially continuous bridge with full-depth
diaphragm at intermediate supports (OBrien et al., 1999)
In the alternative form of partially continuous bridge, illustrated in figure 2.10,

continuity over intermediate supports is provided only by the slab. Thus the in-situ slab alone
is required to resist the complete hogging moment at the intermediate supports. This is
possible due to the fact that members of low structural stiffness (second moment of area) tend
to attract low bending moment. The slab at the support in this form of construction is
particularly flexible and tends to attract a relatively low bending moment. There is concern
among some designers about the integrity of such a joint as it must undergo significant
rotation during the service life of the bridge.
Figure 2.10: Partially continuous bridge with continuity provided only by the slab at
intermediate supports (OBrien et al., 1999)
18
Further, as the main bridge beams rotate at their ends, the joint must move
longitudinally to accommodate this rotation as illustrated in figure 2.11.
Figure 2.11: Joint detail at intermediate support of partially-continuous bridge of the

type illustrated in figure 2.10 (OBrien et al., 1999)
In partially continuous bridges, the precast concrete or steel beams carry all the self
weight of the bridge which generates a bending moment diagram such as that illustrated in
figure 2.12 (a) for a two-span bridge.
Figure 2.12 (a): Bending moment due to selfweight (OBrien et al., 1999)
By the time the imposed traffic loading is applied, the bridge is continuous and the
resulting bending moment diagram is as illustrated in figure 2.12 (b).
Fig. 2.12 (b): Bending moment due to loading applied after bridge has been made
continuous (OBrien et al., 1999)
19
The total bending moment diagram will be a combination of that due to self weight and
other loading. Unfortunately, due to creep, self weight continues to cause deformation in the
bridge after it has been made continuous. At this stage it is resisted by a continuous rather
than a simply supported beam/ slab and it generates a distribution of bending moment more
like that of figure 2.12(b) than figure 2.12(a). This introduces a complexity into the analysis
compounded by a great difficulty in making accurate predictions of creep effects. (OBrien et
al., 1999)
2.3.5 Frame/Box Culvert (Integral Bridge)
Frame or box bridges, such as illustrated in figure 2.13 and 2.14, are more effective at
resisting applied vertical loading than simply supported or continuous beams/slabs.
Figure 2.13: Box culvert (OBrien et al., 1999)
Figure 2.14: Three-span frame (OBrien et al., 1999)
20
This is because the maximum bending moment tends to be less, as can be seen from the
examples of figure 2.15.
(a)
(b)
(c)
Figure 2.15: Typical distributions of bending moment: (a) simply-supported spans; (b)
continuous beams; (c) frames/ box culverts (OBrien et al., 1999)
However, accommodating movements due to temperature changes or creep/shrinkage

can be a problem and, until recently , it was not considered feasible to design frame bridges of
any great length (about 20 m was considered maximum). The effects of deck shortening
relative to the supports is to induce bending in the whole frame as illustrated in figure 2.16 If
some of this shortening is due to creep or shrinkage, there is the usual complexity and
uncertainty associated with such calculations. A further complexity in the analysis of frame
bridges is that, unless the transverse width is relatively small, the structural behaviour is threedimensional. Continuous slab bridges on the other hand, can be analysed using twodimensional models.
21
(a)
(b)
Figure 2.16: Effect of thermal contraction of deck in frame bridge: (a) deflected shape;
(b) distribution of bending moment (OBrien et al., 1999)
The minimal maintenance requirement of frame/ box culvert bridges is their greatest
advantage. There are no joints or bearings as the deck is integral with the piers and abutments.
Given the great upsurge of interest in maintenance and durability in recent years, this lack of
maintenance has resulted in an explosion in the numbers of bridges of this form. Ever longer
spans are being achieved. It is now considered that bridges of this type of 100 m and longer
are possible.
There are two implications for longer frame-type bridges, both relating to longitudinal
movements. If the supports are fully fixed against translation, deck movements in such
bridges will generate enormous stresses. This problem has been overcome by allowing the
supports to slide as illustrated in figure 2.17.
Figure 2.17: Sliding support and run-on slab in frame bridge (OBrien et al., 1999)
22
If the bridge is supported on piles, the axes of the piles are orientated so as to provide
minimum resistance to longitudinal movement. The second implication of longer frame
bridges is that the bridge moves relative to the surrounding ground. To overcome this,
engineers specify 'run-on' slabs as illustrated in the figure which span over loose fill that is
intended to allow the abutments to move. The run-on slab can rotate relative to the bridge
deck but there is no relative translation. Thus, at the ends of the run-on slabs, a joint is
required to facilitate translational movements. Such a joint is remote from the main bridge
structure and, if it does leak, will not lead to deterioration of the bridge itself. (OBrien et al.,
1999)
A precast variation of the frame/box culvert bridge has become particularly popular in
recent years. Precast pretensioned concrete beams have a good record of durability and do not
suffer from the problems associated with grouted post tensioning tendons. These can be used
in combination with in-situ concrete to form a frame bridge as illustrated in figure 2.18
Figure 2.18: Composite precast and in-situ concrete frame bridge (OBrien et al., 1999)
23
2.4
Articulation
In order to understand more about the importance of not having joints and bearings, it is
essential to have the general idea on the matter. Bridge design is often a compromise between
the maintenance implications of providing joints and bearings and the reduction in stresses
which results from the accommodation of deck movements. While the present trend is to
provide ever fewer joints and bearings, the problems of creep, shrinkage and thermal
movement are still very real and no one form of construction is the best for all situations.
The articulation of a bridge is the scheme for accommodating movements due to creep,
shrinkage and thermal effects while keeping the structure stable. While this clearly does not
apply to bridges without joints or bearings, it is a necessary consideration for those which do.
Horizontal forces are caused by braking and traction of vehicles, wind and accidental impact
forces from errant vehicles. Thus, the bridge must have the capacity to resist some relatively
small forces while accommodating movements.
In-situ concrete bridges are generally supported on a finite number of bearings. The
bearings usually allow free rotation but may or may not allow horizontal translation. They are
generally of one of the following three types :
i. fixed - no horizontal translation allowed ;

ii. free sliding - fully free to move horizontally;
iii. guided sliding - free to move horizontally in one direction only.
In many bridges, a combination of the three types of bearing is provided. Two of the
simplest forms of articulation are illustrated in figures 2.19 (a) and (b) where the arrows
indicate the direction in which movements are allowed. For both bridges, A is a fixed bearing
allowing no horizontal movement. To make the structure stable in the horizontal plane, guided
sliding bearings are provided at C and, in the case of the two-span bridge, also at E. These
bearings are designed to resist horizontal forces such as the impact force due to an excessively
high vehicle attempting to pass under the bridge. At the same time they accommodate
longitudinal movements, such as those due to temperature changes. Free sliding bearings are
provided elsewhere to accommodate transverse movements. When bridges are not very wide
24
(less than about 5 m), it may be possible to articulate ignoring transverse movements such as
illustrated in figure 2.19(c).
(a)
(b)
(c)
Figures 2.19: Plan views showing articulation of typical bridges: (a) simply supported
slab; (b) two-span skewed slab; (c) two-span bridge of small width (OBrien et al., 1999)
When bridges are not straight in plan, the orientation of movements tends to radiate
outwards from the fixed bearing. Bearings are generally incapable of resisting an upward
'uplift' force. Further, if unanticipated net uplift occurs, dust and other contaminants are likely
to get into the bearing, considerably shortening its life. Uplift can occur at the acute corners of
skewed bridges such as Band E in figure 2.19( b). Uplift can also occur due to applied loading
in right bridges if the span lengths are significantly different, as illustrated in figure 2.20
25
However, even with no skew and typical span lengths, differential thermal effects can
cause transverse bending which can result in uplift as illustrated in figure 2.21.
Figure 2.20: Uplift of bearings due to traffic loading (OBrien et al., 1999)
Figure 2.21: Uplift of bearing due to transverse loading caused by differential thermal
effects (OBrien et al., 1999)
If this occurs, not only is there a risk of deterioration in the central bearing but, as it is
not taking any load, the two outer bearings must be designed to resist all of the load which
renders the central bearing redundant. Such a situation can be prevented by ensuring that the
reaction at the central bearing due to permanent loading exceeds the uplift force due to
temperature. If this is not possible, it is better to provide two bearings only. (OBrien et al.,
1999)
26
2.5
Bridge Loading
For bridges, it is often necessary to consider phenomena which would normally be

ignored in buildings. For example, effects such as differential settlement of supports
frequently need to be considered by bridge designers while generally being ignored by
designers of building structures. Other types of loading which may occur but which are not
considered here are the effects of shrinkage and creep, exceptional loads (such as snow) and
construction loads.
Another source of loading is earth pressure on substructures. This must be considered

in the context of integral bridges. Three codes of practice can be referred to, namely, the
British Department of Transport standard BD37/ 88 (1988), the draft Eurocode EC1 (1995)
and the American standard AASHTO (1995). Dead and superimposed dead loads consist of
permanent gravity forces due to structural elements and other permanent items such as
parapets and road surfacing. Imposed traffic loads consist of those forces induced by road or
rail vehicles on the bridge. The predominant effect is the vertical gravity loading including the
effect of impact. However, horizontal loading due to braking/traction and centrifugal effects
in curved bridges must also be considered. Where footpaths or cycle tracks have been
provided, the gravity loading due to pedestrians/cyclists can be significant.
Thermal changes can have significant effects, particularly in frame and arch bridges.
Both the British standard and the AASHTO treatments of temperature are somewhat tedious
in that different load 'combinations' must be considered. For example, the AASHTO standard
specifies one combination which includes the effects of temperature, wind and imposed traffic
loading. An alternative, which must also be considered, excludes some thermal and wind
effects but includes a higher traffic loading. The calculation is complicated by the use of
different factors of safety and the specification of different design limits for the different
combinations. For example, the service stresses permitted in prestressed concrete bridges are
higher for the combinations in BD37/88 which include temperature than for combinations
which do not. The draft Eurocode treats temperature in a manner similar to other load types
and applies the same method of combining loads as is used throughout ECl.
27
Differential settlement of supports can induce significant bending in continuous beam
or slab bridges. The draft Eurocode on Geotechnical Design, EC7 (1994), recommends that
the process of soil/structure interaction be taken into consideration for accurate analysis of
problems of this type, i.e. it is recommended that a combined model of the bridge structure
and the supporting soil be used to determine the stresses induced by settlement. No
geotechnical guidance is given in either BD37/88 or AASHTO on how bridges should be
analysed to determine the effect of this phenomenon. The loading due to impact from
collisions with errant vehicles can be quite significant for some bridge elements. The load
specified in the UK has increased dramatically in recent years. Similarly high levels of impact
loading are in use in many European national standards, in AASHTO and in the draft
Eurocode. Vibration is generally only significant in particularly slender bridges. In practice,
this usually only includes pedestrian bridges and long-span road and rail bridges, where the
natural frequency of the bridge is at a level which can be excited by traffic or wind. In
pedestrian bridges, it should be ensured that the natural frequency of the bridge is not close to
that of walking or jogging pedestrians.
In addition to its ability to induce vibration in bridges, wind can induce static
horizontal forces on bridges . The critical load case generally occurs when a train of high
vehicles are present on the bridge resulting in a large vertical projected area. Wind tends not
to be critical for typical road bridges that are relatively wide but can be significant in elevated
railway viaducts when the vertical projected surface area is large relative to the bridge width.
Both the British and the American standards specify a simple conservative design wind
loading intensity which can be safely used in most cases . More accurate (and complex)
methods are also specified for cases where wind has a significant effect.
Prestress is not a load as such but a means by which applied loads are resisted.
However, in indeterminate bridges it is necessary to analyse to determine the effect of
prestress so it is often convenient to treat prestress as a form of loading. The methods used are
very similar to those used to determine the effects of temperature changes. (OBrien et al.,
1999)
28
2.5.1 Dead and Superimposed Dead Loading
For general and building structures, dead or permanent loading is the gravity loading
due to the structure and other items permanently attached to it. In BD37/ 88, there is a
subdivision of this into dead loading and superimposed dead loading. The former is the
gravity loading of all structural elements. It is simply calculated as the product of volume and
material density. For prestressed concrete bridges, it is important to remember that an
overestimate of the dead load can result in excessive stresses due to prestress. Thus dead load
should be estimated as accurately as possible rather than simply rounded up. Superimposed
dead load is the gravity load of non-structural parts of the bridge. Such items are long term
but might be changed during the lifetime of the structure. An example of superimposed dead
load is the weight of the parapet.
There is clearly always going to be a parapet so it is a permanent source of loading.

However, it is probable in many cases that the parapet will need to be replaced during the life
of the bridge and the new parapet could easily be heavier than the original one. Because of
such uncertainty, superimposed dead load tends to be assigned higher factors of safety than
dead 1oad. The most suitable item of superimposed dead load is the road pavement or
surfacing. It is not unusual for road pavements to get progressively thicker over a number of
years as each new surfacing is simply laid on top of the one before it. Thus, such
superimposed dead loading is particularly prone to increases during the bridge lifetime. For
this reason, a particularly high load factor is applied to road pavement.
Bridges are unusual among structures in that a high proportion of the total loading is
attributable to dead and superimposed dead load. This is particularly true of long-span
bridges. In such cases, steel or aluminium decks can become economically viable due to their
high strength-to-weight ratio. For shorter spans, concrete or composite steel beams with
concrete slabs are the usual materials. In some cases, lightweight concrete has been
successfully used in order to reduce the dead load.
In addition to its ability to induce vibration in bridges, wind can induce static
horizontal forces on bridges. The critical load case generally occurs when a train of high
vehicles are present on the bridge resulting in a large vertical projected area. Wind tends not
29
to be critica l for typical road bridges that are relatively wide but can be significant in elevated
railway viaducts when the vertical projected surface area is large relative to the bridge width.
Both the British and the American standards specify a simple conservative design wind
loading intensity which can be safely used in most cases. More accurate (and complex)
methods are also spicified for cases where wind has a significant effect. (OBrien et al., 1999)
2.5.2 Imposed Traffic Loading
Bridge traffic can be vehicular, rail or pedestrian cycle or indeed any combination of
these. Vehicular and rail traffic are considered in subsections below. While pedestrian/cycle
traffic loading on bridges is not difficult to calculate, its importance should not be
underestimated. Bridge codes commonly specify a basic intensity for pedestrian loading (e.g.
5 kN/m2 in the draft Eurocode and the British standard and 4 kN/m2 in the American code).
When a structural element supports both pedestrian and traffic loading, a reduced intensity is
allowed by some codes to reflect the reduced probability of both traffic and pedestrian loading
reaching extreme values simultaneously. Most codes allow a reduction for long footpaths.
2.5.3 Imposed Loading Due to Road Traffic
While some truck-weighing campaigns have been carried out in the past, there has
been a scar city of good unbiased data on road traffic loading until recent years. Bridge traffic
loading is often governed by trucks whose weights are substantially in excess of the legal
maximum. In the past, sampling was carried out by taking trucks from the traffic stream and
weighing them statically on weighbridges. There are two problems with this as a means of
collecting statistics on truck weights. In the first place, the quantity of data collected is
relatively small but, more importantly, there tends to be a bias as drivers of illegally
30
overloaded trucks quickly learn that weighing is taking place and take steps to avoid that point
on the road.
In recent years the situation has improved considerably with the advent of weigh-inmotion (WIM) technology which allows all trucks passing a sensor to be weighed while they
travel at full highway speed. WIM technology has resulted in a great increase in the
availability of truck weight statistics and codes of practice are being revised to reflect the new
data.
Bridge traffic loading is applied to notional lanes which are independent of the actual
lanes delineated on the road. In the Eurocode, the road width is divided into a number of
notional lanes, each 3 m wide. The outstanding road width, between kerbs, after removing
these lanes, is known as the ' remaining area'. (OBrien et al., 1999)
2.5.4 Thermal Loading
As integral bridge is concerned, there are considerable durability advantages in

removing joints and bearings, but their removal does affect the bridge behaviour. Specifically,
expansion and contraction of the deck is restrained with the result that additional stresses are
induced which must be resisted by the bridge structure. The most obvious cause of expansion
or contraction in bridges of all forms is temperature change.
There are two thermal effects which can induce stresses in bridges. The first is a
uniform temperature change which results in an axial expansion or contraction. If restrained,
such as in an arch or a frame bridge, this can generate significant axial force, bending moment
and shear. The second effect is that due to differential changes in temperature. If the top of a
beam heats up relative to the bottom, it tends to bend while if it is restrained from doing so,
bending moment and shear force are generated.
If a beam is on a sliding bearing as illustrated in figure 2.22 (a) and the temperature is
reduced by T, it will contract freely. A (negative) strain will occur of magnitude (T)
31
where is the coefficient of thermal expansion (strain per unit change in temperature). The
beam then contracts by (T) l where l is its length. However, no stresses are generated as no
restraint is offered to the contraction . As there is no stress, there can be no tendency to crack.
If, on the other hand, the beam is fixed at both ends as illustrated in figure 2.22 (b), and its
temperature is reduced by T, then there will be no strain. There cannot be any strain as the
beam is totally restrained against contraction. This total restraint generates a stress of
magnitude E (T), where E is the elastic modulus. The stress is manifested in a tendency to
crack.
Figure 2.22 (a): Beam on sliding bearing (OBrien et al., 1999)
Figure 2.22 (b): Beam fixed at both ends (OBrien et al., 1999)
Uniform changes in temperature result from periods of hot or cold weather in which
the entire depth of the deck undergoes an increase or decrease in temperature. Both the draft
Eurocode and the British standard specify contour plots of maximum and minimum ambient
temperature which can be used to determine the range of temperature for a particular bridge
site. The difference between ambient temperature and the effective temperature within a
bridge depends on the thickness of surfacing and on the form of construction (whether solid
slab, beam and slab, etc.). The American approach is much simpler. In 'moderate' climates,
metal bridges must be designed for temperatures in the range - 18C to 49 C and concrete
bridges for temperatures in the range - 12 C to 27 C. Different figures are specified for 'cold'
climates.
32
It is important in bridge construction to establish a baseline for the calculation of
uniform temperature effects, i.e. the temperature of the bridge at the time of construction. It is
possible to control this baseline by specifying the permissible range of temperature in the
structure at the time of completion of the structural form. Completion of the structural form
could be the process of setting the bearings or the making of a frame bridge integral. In
concrete bridges, high early temperatures can result from the hydration of cement, particularly
for concrete with high cement contents. Resulting stresses in the period after construction will
tend to be relieved by creep although little reliable guidance is available on how this might be
allowed for in design. Unlike in-situ concrete bridges, those made from precast concrete or
steel will have temperatures closer to ambient during construction. The AASHTO code
specifies a baseline temperature equal to the mean ambient in the day preceding completion of
the bridge. The British Standard and the draft Eurocode specify no baseline.
Integral bridges undergo repeated expansions and contractions due to daily or seasonal
temperature fluctuations. After some time, this causes the backfill behind the abutments to
compact to an equilibrium density. In such cases, the baseline temperature is clearly a mean
temperature which relates to the density of the adjacent soil. In addition to uniform changes in
temperature, bridges are subjected to differential temperature changes on a daily basis, such as
in the morning when the sun shines on the top of the bridge heating it up faster than the
interior. The reverse effect tends to take place in the evening when the deck is warm in the
middle but is cooling down at the top and bottom surfaces. Two distributions of differential
temperature are specified in some codes, one corresponding to the heating-up period and one
corresponding to the cooling-down period. These distributions can be resolved into axial,
bending and residual effects.
As for uniform changes in temperature, the baseline temperature distribution is

important, i.e. that distribution which exists when the structural material first sets. However,
no such distribution is typically specified in codes, the implication being that the distributions
specified represent the differences between the baseline and the expected extremes.
Transverse temperature differences can occur when one face of a superstructure is subjected
to direct sun while the opposite side is in the shade. This effect can be particularly significant
when the depth of the superstructure is great.
33
Cracking of reinforced concrete members reduces the effective cross sectional area
and second moment of area. If cracking is ignored, the magnitude of the resulting thermal
stresses can be significantly overestimated. The effects of both uniform and differential
temperature changes can be determined using the method of 'equivalent loads'. A distribution
of stress is calculated corresponding to the specified change in temperature. This is resolved
into axial, bending and residual distributions. The corresponding forces and moments are then
readily calculated. (OBrien et al., 1999)
34
CHAPTER 3
RESEARCH METHODOLOGY
3.1
Introduction
Generally, the steps in this study can be divided into five different parts. The first step
is the literature review on how to model integral bridge, the effects of temperature on integral
bridge, and how to use finite element software, LUSAS. After the literature review, the
second step is data collection which consists of the data on the model to be developed,
Malaysian temperature gradient, and also the current status of the usage of integral bridge in
Malaysia. The third step is to develop a model of the integral bridge and to verify the
developed model by consulting with the experts. The verified model will then be manipulated
to study the significance of thermal effects on integral bridge in Malaysian weather and
temperature. With the intention of achieving research objectives, it is important to implement
the right approach in order to address the problems. This chapter will explain the details on
research method that will be adopted along the study. The flow chart of this study is shown in
figure 3.1.
35
Problem
Identification
Data
Collection
Data from previous

research
Consultation with
LUSAS users and
Finite Element
specialists
Adaptation of real
integral bridge design
Development of the
model using LUSAS
Verification of the
model
Verified model is used

to analyze and study
the effect of
temperature for various
length of integral
bridge
Findings and
conclusion
Figure 3.1: Methodology flow chart
36
3.2
Problem Identification
It is understandable that integral bridge offers a great minimization of costs compared

to conventional type of bridge. However, expansion and contraction of the deck is restrained
with the result that additional stresses are induced. This additional stresses must be resisted by
the bridge structure. Temperature is one of the most dominant factors in the bridges
expansion and contraction but is it significant in Malaysian condition? Will the thermal
effects only be significant at certain length of the bridge? How much is the changes in the
prestressed force? The questions to these answers are the main objective of this study.
3.3
Data Collection
The data collection sources are mainly from previous research, discussion with
experienced LUSAS users and also adaptation of the real design of integral bridge.
3.3.1 Previous Research
This research is inspired by the previous study conducted by Charles D. Newhouse et

al entitled Modelling Early-Age Bridge Restraint Moments: Creep, Shrinkage, and
Temperature Effects. A lot about the temperature effects is learned from this research. The
research mainly consists of experimental study, monitoring the early age restraint moments
that develop in a two-span continuous system made of full-depth precast concrete bulb tee
girders and comparison of the restraint moments observed to the predicted restraint moments
using the RMCalc program. They also proposed a simplified model to predict the restraint
moments considering the thermal effects. The overall elevation of the test setup is shown in
the following figure 3.2 and figure 3.3;
37
Figure 3.2: Overall elevation of the Charles D. Newhouse research test setup (Newhouse
C. D. et al., 2008)
Figure 3.3: Diaphragm details (Newhouse C. D. et al., 2008)
There were two tests monitored in order to achieve the purpose of the experimental
study; that is to compare the restraint moments that developed during the early ages of
continuity to the predicted restraint moments using RMCalc program. The two tests were
monitored for the development of restraint moments in a continuous system from the time the
deck was cast until the deck reached its design compressive strength. Full-depth specimens
and full-width deck sections were used in the testing to simulate actual conditions. The
changes in end reactions on the two span systems were carefully monitored and recorded
throughout the testing. The restraint moments were determined by multiplying the change in
the end reaction by the span length. Finally, a simplified model was developed to predict the
restraint moments considering early age thermal effects. This model is presented as an
alternate way to predict restraint moments.
38
3.3.2 Adaptation of Real Integral Bridge Design
A real integral bridge design of Pantai Timur Second Phase Highway Project
designed by Ir. Mohamed Salleh Yassin is adapted as the basis in developing the integral
bridge model in LUSAS software.
The real design of the integral bridge is then simplified to a much simpler design,
therefore it is easier to be modelled. It is important to emphasize that the main objective of the
development of the model is to study the effect of temperature on the prestressed integral
bridge beam. Hence, a simple design of an integral bridge would suffice but at the same time,
the attributes of a real integral bridge design is needed.
3.3.2.1 Model Layout
The following figure shows a simple sketch of the integral bridge to be modelled.
Assumptions and conditions are applied in the development of the model. In this study, the
analysis is done for integral bridge with spans 20m, 30m and 40m.
Figure 3.4: Longitudinal integral beam
In order to make sure that the results for all spans are not too affected by the
geometrical difference, the beam geometry is designed so that the safety factor for each
integral bridge span is about the same. The safety factors for each bridge span is shown in
table 2;
39
Table 3.1: Safety factor calculated according to each bridge spans Z1 and Z2
actual (mm3)
required (mm3)
Safety factor
length
z1
z2
z1
z2
z1(actual)/z1(required)
z2(actual)/z2(required)
20
147.05
150.9
123.21
107.13
1.19
1.41
30
293.21
294.48
244.1
212.23
1.20
1.39
40
501.56
504.79
417.84
363.28
1.20
1.39
Table 3.2: General details for beam span 20m, 30m and 40m
Beam span (m)
20
30
40
6000
7200
11200
No of strand
48
60
90
No of Tendon
Type of strand
7-wire standard strand
7-wire standard
7-wire standard
12.9 mm
strand 12.9 mm
strand 12.9 mm
1500
1800
2800
Prestressing force (kN)
P at each tendon (kN)
3.3.2.2 Support Connection Detail
Figure 3.5: Integral beam-to abutment connection (Connal J., 2003)
40
Figure 3.6: Integral beam-to abutment connection detail. (Connal J., 2003)
3.4
LUSAS Structural Modelling
LUSAS Modeller is an associative feature-based modelling system. The model

geometry is entered in terms of geometry features which are sub-divided into finite elements
in order to perform an analysis. Increasing the density of the mesh will usually result in an
increase in accuracy of the solution, but with a corresponding increase in solution time and
disk space required. The geometry features form a hierarchy, that is volumes which are
enclosed by surfaces, which are made up of lines, which are defined by end points
A LUSAS model is graphically represented by geometry features (points, lines,
surfaces, volumes) which are assigned attributes (materials, thicknesses, loading, supports,
mesh, etc.). In developing the finite element model, the following general guidelines are
followed:
a. Understand the structural behaviour of the problem to be solved which includes
physical behaviour, loading type and magnitude, and boundary conditions
41
b. Understand the behaviour and limitations of applicable theory of various
elements available for use. This will help the choosing of proper types of
elements to match as close as possible to the physical behaviour of the problem
c. Understand the programs option and limitations
d. A good practice is to; work with experienced people and search for reference
e. Avoid 3D because most problems can be modelled in 2D: Plane truss, plane
frame, plane stress, plane strain, axis symmetric, plate bending
f. Understand how various elements behave in various situations
g. Understand the physics of the problems well enough to make an intelligent
choice of elements and mesh
h. Start with a simple model and gradually refine them to a more complicated
one.
i. Anticipate the results and know the goal
j. Check the model
k. Check the results and typical post processor results
3.4.1 Assumptions
A few assumptions are considered in this model. They are;
1) The model is built using only line element.

2) The HA load is only consider based on the spacing between the beam, that is, 1900
mm.
3) The height of abutment is fixed to 6000 mm for all models.
4) Beam cross-section is the same along the span
5) The post-tensioned beam is designed for 4 tendons, but only 1 tendon is considered in
the model
6) Only 1 beam is considered in the integral bridge model
7) The abutment is designed for 1m depth
42
3.4.2 Geometry Definition
The integral bridge with a prestressed beam basically can be modelled with
combination of lines. The lines are grouped as integral bridge beam and tendon profile. For
the integral bridge line, straight line is used to draw the bridge beam connecting to the bridge
abutment. Meanwhile, spline curve is used to visualise the parabolic tendon profile line.
These lines are easily drawn using the coordinates system in the software. It is shown in the
following figure 3.7 and figure 3.8:
Figure 3.7: Integral bridge beam line
Figure 3.8: Tendon profile line
3.4.3 Attributes Definition
Attributes are used to describe the properties of the model. Attributes are assigned to
geometry features and are not lost when the geometry is edited, or the model is re-meshed.
Attribute assignments are inherited when geometry features are copied and are retained when
43
geometry features are moved. The attribute types can be categorized as general attributes and
specific attributes. To study temperature effects, only general attributes is used. The general
attributes are:
i.
Mesh
: describes the element type and discretisation on the geometry.
ii.
Geometric
: specifies any relevant geometrical information that is not inherent in

the feature geometry, for example section properties or thickness.
iii.
Material
: defines the behaviour of the element material, including linear,

plasticity, creep and damage effects.
iv.
Support
: specifies how the structure is restrained. Applicable to structural, pore

water and thermal analyses.
v.
Loading
: specifies how the structure is loaded
3.4.3.1 Meshing Attribute
Model for the integral bridge in this study uses beam elements. In LUSAS, beam
elements are used to model plane and space frame structures. LUSAS incorporates a variety
of thin and thick beams in both 2 and 3-dimensions. In addition, specialised beam elements
for modelling grillage or eccentrically ribbed plate structures are also available. LUSAS beam
elements may be either straight or curved and may model axial force, bending and torsional
behaviour.
In this case, 2 dimensional thick beam structural elements with linear interpolation
order is used. For the prestressed beam, the number of divisions used vary from 20, 30 to 40
depending on the length of the span. The more number of divisions, the more accurate will the
model be. For instance, when analysing 30m span integral bridge, 30 number of elements are
used. At the same time, the number of division for the abutment is fixed at 6 for all three span
lengths. The following figures show the meshing assignment interface in LUSAS and also
active meshing applied to the beam and abutments respectively;
44
Figure 3.9: Line mesh assignment interface
Figure 3.10: Active mesh applied to beam and abutments
45
3.4.3.2 Geometric Attribute
There are 3 bridge spans to be analyzed. Therefore, there are 3 different beam designs
with different section properties. The section properties for each beam designs are calculated
using LUSAS. After the section properties for specific beam design is calculated, the data
were then stored in the section library database of LUSAS. Geometric assignment for the
specific bridge span length can be done by simply selecting the stored data from the section
library database. Figure below shows the arbitrary section property calculator interface used
in LUSAS to calculate and store the section properties data.
Figure 3.11: Arbitrary Section Property Calculator Interface
Geometric properties which have not been defined by the feature geometry are
assigned using geometric attributes. The properties required are element dependent and are
defined for an element family such as bars, beams or shells etc. In this case, the attribute used
is beam elements and it is then assigned to the required line.
Figure 3.12, figure 3.13 and figure 3.14 shows the cross section and the dimensions
for 20m, 30m and 40m span length respectively. Meanwhile, table 3.3 shows the section
properties of the spans.
46
Table 3.3: Section properties for 20m, 30m and 40m integral bridge span
Beam span (m):
20
30
40
0.583 x 106
0.74 x 106
0.943 x 106
Y1
517
822
1088
504
818
1082
76.04x109
241.02x109
545.94x109
Z1
147.05 x 106
293.21 x 106
501.56 x 106
Z2
150.90 x 106
294.48 x 106
504.79 x 106
Cross-sectional area (mm2)

Centroid (mm):
Y2
Moment Inertia (mm4)

Section Moduli (mm3):
Figure 3.12: 20m integral bridge span cross-section and dimensions
47
48
49
3.4.3.3 Material Attribute
Different definitions of materials are used for the integral beam and also the
abutments. For the beam, the material used is long term concrete base on BS5400 with
concrete strength of 50 N/mm2. Meanwhile, for the abutment, the material used is long term
concrete base on BS8110 with concrete strength of 50 N/mm2. Both of these materials are
isotropic material and allow thermal expansion for analysis.
Table 3.6 below shows the material properties for both abutment and beam. All units
are in kN, mm, kt, s, C. Figure 3.15 shows the material assignment interface in LUSAS.
Table 3.4: Material properties.

Long Term Concrete
Long Term Concrete
BS5400
BS8110
Concrete Grade
C50
C50
Young's Modulus
17
15
Poisson's Ratio
0.2
0.2
Mass Density
2.4 E-12
2.4 E-12
0.012 E-3
0.01 E-3
Material
Coefficient of Thermal
Expansion
50
Figure 3.15: Material assignment interface
3.4.3.4 Support Attribute
Support conditions describe the way in which the model is restrained. A support
attribute contains information about the restraints to be applied to each degree of freedom.
There are three valid support conditions:
i.
Free
: the degree of freedom is completely free to move. This is the default.
ii.
Fixed : the degree of freedom is completely restrained from movement.
51
iii.
Spring Stiffness : the degree of freedom is subjected to a specified spring stiffness.

Spring stiffness values can be applied uniformly to All nodes meshed on the assigned
feature or their values may vary over a feature by applying a variation. Alternatively,
per unit length or per unit area values can be applied
For this study, the supports used are roller and spring support. For roller support,
translation in y, z and rotation in x are set to be fixed. The rest are set to be free. Figure 3.16
shows the structural support assignment for roller.
Figure 3.16: Structural support setting for roller
Spring support is used to represent the soil pressure behind the abutment. The spring
stiffness is set at translation x with varying values according to the span length analysis while
the spring stiffness distribution is set for stiffness per unit length. Figure 3.17 shows the
structural support assignment for spring stiffness. Figure 3.18 shows the visual of roller
support and spring stiffness support in LUSAS.
52
Figure 3.17: Structural support setting for spring stiffness
Spring stiffness support
Roller support
Figure 3.18: Visual of roller support and spring stiffness support in LUSAS
53
3.4.3.4.1 Example Calculation for Spring Stiffness
Dry density of the backfill, d
= 1600 kg/m3
Coefficient of thermal expansion, = 12 x 10-6 per oC

Specific gravity, Gs
= 2.65
Increased temperature, T
= 20 oC
Horizontal stress, p
= 50 kN/m2
It is assume that the density of the backfill reaches an equilibrium value 20% in excess of that
specified.
d = 1.2(1600) = 1920 kg/m3

d =
e = Gs w/ d -1
= (2.65)(1000)/(1920) 1
=0.38
= T L/4
= (12 x 10-6)(20)(20/4)
= 1.2 x 10-3 m
= 2/3H
= 2(1.2 x 10-3)/3(6)
= 1.33 x 10-4
Es = 150 x 103
= 150 x 103
= 219496.572 kN/m2
54
khorz =
=
= 136 482.623 kN/m/m2
Table 3.2 shows the value of spring stiffness for all span length.
Table 3.5: Values of spring stiffness.
Span Length (m)
khorz (kN/m/m2)
20
136 482.623
30
116 048.854
40
103 434.486
3.4.3.5 Loading Attribute
The loading considered in this study are self weight, HA bridge loading and
temperature loading. In LUSAS, self weight can be defined by assigning the linear gravity
acceleration, 9.81 m/s to y-direction.
For HA load, the following is the example data that used in this study;
Table 3.6: HA loading details.
Loading Code
BD37/01
Loaded Length(m)
30
Notional Lane Width (m)
2.88
Skew Angle deg)
55
As stated earlier, there are two types of temperature effects. They are the uniform
temperature change, and differential temperature change. From these two temperature effects,
the cases are then elaborated into four conditions as shown in figure 3.19.
TTop
TBottom
(a)
TTop
TBottom
(b)
(c)
(d)
Figure 3.19: Four conditions of temperature effects;
(a) Temperature gradient (TT > TB), (b) Temperature gradient (TT < TB),
(c) Uniform temperature increment, (d) Uniform temperature decrement
56
3.4.4 Prestress Definition to BS5400
The single tendon prestress wizard allows loading due to post tensioning to be
generated and hence an analysis to determine the stresses in post tensioned concrete structures
can be computed.
Computation of tendon forces due to post tensioning can be carried out in accordance
with BS5400, AASHTO-LRFD or Eurocode. Direct import of tendon forces may also be
defined via an Excel spreadsheet. In this study, the prestress force is defined in accordance to
BS5400.
As stated before, the concrete beam is modelled as beam elements while the tendon
profile is defined as a spline or combined line. Note that sufficient points must be used in the
line definition to accurately represent the tendon profile as a series of straight lines. The
tendon loads are applied using either beam element loads.
The assignment interface of prestress force for long term is shown by figure. Figure
shows the example of prestress assigned to the tendon:
Figure 3.20: Single tendon prestress assignment interface according to BS5400
57
Figure 3.21: Visual of assigned prestress at tendon
3.5
Prestress Change Determination by Trial and Error Method
After the model is developed, it is then used to determine the prestress change in the
integral bridge beam. A trial and error method is used. However the method only determined
the value of the prestress change, not the orientation of the change. To determine whether the
value of prestress change is a loss or an addition to the initial prestress force, the behaviour of
the beam due to temperature effect is looked into.
As stated earlier, there are two types of temperature effects. They are the temperature
gradient effect and uniform temperature effect. When the beam is affected by temperature
gradient loading, it will tend to bend. Whether it bends upwards or downwards depends on
which is hotter, the extreme top fibre or the extreme bottom fibre. If the temperature at the top
beam is higher than the temperature at the bottom of the beam, then it will tend to bend
upwards. This will result in relaxation of the tendon that means there will be a loss in the
prestress force.
As for uniform temperature change, the beam will tend to expand or contracts
depending on whether it is temperature increment or decrement. When the case is uniform
58
temperature increment, then the beam will tend to expand, thus causing an additional
prestressing force to be induced in the tendon.
After it is clear on how will the beam reacts to different cases of the temperature
effects, the trial and error method to determine the value of prestress change is then
conducted. The method is simple. Basically there will be 3 stages;
i.
The first stage is to determine the deflection due to self weight, live load (HA load)
and long term prestressing force.
ii.
The second stage will be to determine the deflection due to self weight, live load (HA
load), long term prestressing force and also the deflection due to the effect of
temperature.
iii.
After the deflection in the first and second stage is known, the third stage is to
determine the same deflection as the second stage, but to use the load case in the first
stage. The variant will be the prestressing force. This is where the trial and error
method comes into effect.
As an example, take an integral bridge with a span of 20m, with an initial prestressing
force of 1500kN. The temperature gradient where the temperature at the extreme top fibre is
larger than the temperature at the extreme bottom fibre is 30 C. What is the value of prestress
decrement due to the effect of temperature?
i.
First stage :
Self weight + HA loading + Prestress force (1500kN)

Deflection = 29.07mm
ii.
Second stage :
Self weight + HA loading + Prestress force (1500kN) + Temperature Gradient 30 C

59
iii.
Third stage:
Self weight + HA loading + Prestress force ( ? kN)

Prestress force is adjusted so that the deflection value due to selfweight and short term
prestress loss is about the same with the deflection value due to selfweight, short term
prestress loss and temperature loading. Table shows the trial and error method.
Table 3.7: Trial and error method

Trial
Prestress force (kN)
Deflection (mm)
2000
27.49
3000
24.41
2500
25.94
2550
25.79
From the trial and error method, it is determined that 2550kN of prestressing force is
needed in order to achieve the deflection of 25.79mm. Therefore, the prestress change at
tendon for 20m span integral bridge under the effect of temperature gradient of 30 C is
2550kN 1500kN = 1050kN. For temperature gradient where the temperature at the extreme
top fibre is larger than the temperature at extreme bottom fibre, there will be a loss of
prestress.
Effective prestress force = 1500kN 1050kN = 450kN
3.6
Comparison of LUSAS and Hand Calculation Trial and Error Method
The trial and error method is a simple and logic method to determine the value of the
prestress change in the tendon due to the effect of temperature loading. In this study, the trial
and error method is based on LUSAS. In order to double check the LUSAS model, the value
60
of prestress change determined by trial and error based on LUSAS is compared to the trial and
error method by hand calculation.
For the sake of simplicity, the prestress change determination is done for simply
supported beam instead of the integral bridge beam. The purpose of the comparison is to show
that the data calculated from LUSAS model did not differ too much from the hand calculation
method. Therefore, the LUSAS model can be verified and applicable for the study of
temperature effect on the prestress integral bridge beam.
3.6.1 Example Calculation for Hand Calculation Method
The following is an example of the hand calculation trial and error method. The trial
and error method is based on the deflection calculation. Therefore, the hand calculation
basically consists of deflection calculations.
Specification:
TT
TB
a. Simply supported bridge with 20m span
b. Section depth, h
= 1021mm
c. Prestress or jacking force
= 6000kN
d. Eccentricity at mid-span,emax = 400mm

e. Eccentricity at support, emin = -67mm
f. Loss of prestress
= 29%
g. Area, A
= 582923 mm2
h. Moment of inertia, I
= 76 x 109 mm4
i. Modulus of elasticity, Ec
= 30 kN/mm2
j. Characteristic strength, fcu
= 50 kN/mm2
61
k. Modulus of elasticity, Es
= 195 kN/mm2
l. Concrete creep coefficient, = 1

m. T
= TT TB
= 30 C
n. Coef. of thermal expansion = 12 x 10-6 per C
Loadcase considered = Selfweight of beam + HA loading (live load) + Temperature loading
Therefore;
Deflection of beam, b= deflection due to prestress, p + deflection due to selfweight,sw
+ Deflection due to HA loading, HA
+ Deflection due to temperature loading, T
Assumptions:
i.
All deflection calculations are long term deflection
ii.
Temperature varies linearly
iii.
No temperature 'discontinuity' between member surface and air.
Deflection due to selfweight,sw

Curvature, 1/r
= M/EcI
Selfweight,w = 14.57 kN/m

At midspan: M = wL2/8 = 729kNm
1/r
= 729 x 106 / (30 x 103 x 76 x 109)

= 0.319 x 10-6 mm-1
sw
= L2/9.6 (1/r)
= (200002/9.6) x 0.319 x 10-6
= 13.31 mm downwards
Long term deflection = (1 + ) x short term deflection

= (1 + 1) x 13.31
62
Deflection due to HA loading,HA
Curvature, 1/r = M/EcI
Live load,w
= 29.79 kN/m
At midspan: M = wL2/8 = 1490 kNm

1/r
= 1490 x 106 / (30 x 103 x 76 x 109)

= 0.653 x 10-6 mm-1
HA
= L2/9.6 (1/r)
= (200002/9.6) x 0.653 x 10-6
Long term deflection = (1 + ) x short term deflection

= (1 + 1) x 27.21
Deflection due to prestress, p

Curvature, 1/r = Pe/Ec I ( + (1 + )/2)
At midspan: =[ (6000 x 103 x 400)/ (30 x 103 x 76 x 109)] x (0.71 + 0.856)
= 1.7 x 10-6 mm-1
At support:
=[ (6000 x 103 x -67)/ (30 x 103 x 76 x 109)] x (0.71 + 0.856)

= -0.3 x 10-6 mm-1
= L2/9.6 [(1/r1) + (1/5)(1/r2)]

= (200002/9.6) x [1.7 + (0.2 x -0.276)] x 10-6
= 66.48 mm upwards
63
Deflection due to temperature loading, T
Figure 3.22: Estimation of the effects of unequal extreme fibre temperatures by the
flexibility method
= curvature = 1/R
M/I
= E/R,
M = Ei
= EiT/h
At midspan:
= (30 x 76 x 109 x 12 x 10-6 x 30) / 1021

= 803.92kNm
1/r
= 803.92 x 106 / (30 x 103 x 76 x 109)

= 0.352 x 10-6 mm-1
sw
= L2/9.6 (1/r)
= (200002/9.6) x 0.352 x 10-6
= 14.7 mm upwards
64
Therefore;
Deflection of beam, b = deflection due to prestress, p + deflection due to selfweight,sw
+ deflection due to HA loading, HA
+ deflection due to temperature loading, T
= 66.48 mm upwards + 26.618 mm downwards
+ 54.412 mm downwards + 14.7 mm upwards
= 0.15 mm upwards
Table 3.10 below shows the comparison of prestress reduction percentage determined using
LUSAS and also by hand calculation. From the table we can summarize that there is not much
of a difference between the 2 methods. Therefore, for analysis, only the data from LUSAS is
used.
Table 3.8: Comparison of the reduction percentage between LUSAS and hand calculation
LUSAS
Hand calculation
Temperature gradient,
Initial
TT > TB
Prestress
Prestress
Reduction
Prestress
Reduction
( C)
(kN)
reduction
percentage
reduction
percentage
(kN)
(%)
(kN)
(%)
3.7
10
6000
485
8.1
420
7.0
20
6000
970
16.2
840
14.0
30
6000
1450
24.2
1260
21.0
40
6000
1940
32.3
1685
28.1
Analysis and Result Interpretation
After the model is verified, now we know that the results are valid. Thus, the model
can be run and analyzed to study the effect of temperature effects on prestress integral bridge
beam.
CHAPTER IV
RESULT AND DISCUSSION
4.1
Introduction
Finite element analysis using LUSAS software is done. The LUSAS model is verified
by comparing it with the hand calculation method. Thus, an integral bridge finite element
model is developed. The next step is to manipulate the data obtained from the analysis to
study the effect of temperature on prestressed integral bridge beam. In this chapter, the results
is discussed and related to the objectives of this study.
4.2
Analysis Outline
In order to study the effect of temperature on prestressed integral bridge beam, the
change in the prestressing force is looked into. The behavior of the beam due to the four cases
of temperature effects can determine whether the prestress is at loss or addition. These
changes in prestressing force are then related to the bridge span variables. We would really
want to know whether the bridge span will give a big difference in the prestress change. On
66
top of that, the temperature change variable to the prestress change is also studied to
determine whether temperature change at certain values can be neglected.
4.3
Comparison of Deflection
Displacement or deflection is an important aspect in designing a prestress beam.

Tolerance will allow the structure to displace at certain extent. However, excessive vertical
displacement can lead to members cracking and the worst case scenario is structural failure. In
this analysis, only deflection at middle span is considered. Tables and figures of 4.1, 4.2, 4.3
and 4.4 showed the vertical displacement integral bridge under self weight, HA loading,
prestressing force and temperature effects for 20, 30 and 40 meter span respectively.
4.3.1 Case 1: Temperature gradient (TT > TB)
TTop
TBottom
Figure 4.1: Temperature gradient (TT>TB) case for deflection comparison

Table 4.1.1: Mid-span deflection comparison for 20m span integral bridge and simply
supported bridge due to temperature gradient (TT>TB)
TB to TT ( c )
Temp. Gradient ( c )
IB deflection (mm)
30 to 40
20 to 40
10 to 40
0 to 40
10
20
30
40
-27.98
-26.89
-25.79
-24.70
SSB deflection (mm)

-69.85
-63.85
-57.85
-51.85
67
TB to TT ( c )
30 to 40
20 to 40
10 to 40
0 to 40
10
20
30
40
IB deflection (mm)
-58.81
-54.36
-49.91
-45.46
SSB deflection (mm)

-115.95
-102.45
-88.95
-75.45
TB to TT ( c )
30 to 40
20 to 40
10 to 40
0 to 40
10
20
30
40
IB deflection (mm)
-109.50
-98.41
-87.34
-76.28
SSB deflection (mm)

-176.08
-152.08
-128.08
-104.08
Figure 4.2: Graph of mid span deflection versus temperature gradient (TT>TB) for IB and SSB
68
4.3.2 Case 2: Temperature Gradient (TT < TB)
TTop
TBottom
Figure 4.3: Temperature gradient (TT<TB) case for deflection comparison

supported bridge due to temperature gradient (TT<TB)
TB to TT ( c )
40-30
40-20
40-10
40-0
10
20
30
40
IB deflection (mm)
-30.17
-31.26
-32.35
-33.45
SSB deflection (mm)

-81.85
-87.85
-93.85
-99.85
TB to TT ( c )
40-30
40-20
40-10
40-0
10
20
30
40
IB deflection (mm)
-67.71
-72.16
-76.61
-81.06
SSB deflection (mm)

-142.95
-156.45
-169.95
-183.45
TB to TT ( c )
40-30
40-20
40-10
40-0
10
20
30
40
IB deflection (mm)
-131.60
-142.70
-153.70
-164.80
SSB deflection (mm)

-224.08
-248.08
-272.08
-296.08
69
Figure 4.4: Graph of mid span deflection versus temperature gradient (TT<TB) for IB and SSB
4.3.3 Case 3: Uniform Temperature Increment
Figure 4.5: Uniform temperature increment case for deflection comparison
70
supported bridge due to uniform temperature increment
Temp. change ( c )
30 to 40
20 to 40
10 to 40
0 to 40
10
20
30
40
IB deflection (mm)
-27.35
-25.62
-23.89
-22.17
SSB deflection (mm)

-75.85
-75.85
-75.85
-75.85
Temp. change ( c )
30 to 40
20 to 40
10 to 40
0 to 40
10
20
30
40
IB deflection (mm)
-60.19
-57.11
-54.03
-50.96
SSB deflection (mm)

-129.45
-129.45
-129.45
-129.45
Temp. change ( c )
30 to 40
20 to 40
10 to 40
0 to 40
10
20
30
40
IB deflection (mm)
-116.20
-111.90
-107.50
-103.20
SSB deflection (mm)

-200.08
-200.08
-200.08
-200.08
71
Figure 4.6: Graph of mid span deflection versus uniform temperature increment for IB and
SSB
4.3.4 Case 4: Uniform Temperature Decrement
Figure 4.7: Uniform temperature decrement case for deflection comparison
72
supported bridge due to uniform temperature decrement
Temp. change ( c )
40-30
40-20
40-10
40-0
10
20
30
40
IB deflection (mm)
-30.80
-32.52
-34.25
-35.98
SSB deflection (mm)

-75.85
-75.85
-75.85
-75.85
Temp. change ( c )
40-30
40-20
40-10
40-0
10
20
30
40
IB deflection (mm)
-66.34
-69.41
-72.49
-75.56
SSB deflection (mm)

-129.45
-129.45
-129.45
-129.45
Temp. change ( c )
40-30
40-20
40-10
40-0
10
20
30
40
IB deflection (mm)
-124.90
-129.20
-133.50
-137.90
SSB deflection (mm)

-200.08
-200.08
-200.08
-200.08
73
Figure 4.8: Graph of mid span deflection versus uniform temperature decrement for IB and
SSB
4.4
Prestress Change in Prestressed Integral Bridge Beam
When a beam is bended whether upwards or downwards, or when a beam is expanding

or contracting, there will be an additional stress or loss of stress to the initial prestress force in
the tendon. These are the so called prestress change. The prestress change differs for the four
cases of temperature effects. Tables and figures of 4.5, 4.6, 4.7 and 4.8 showed the prestress
change in the tendon for integral bridge under self weight, HA loading, prestressing force and
temperature effects for 20, 30 and 40 meter span respectively.
74
4.4.1 Case 1: Temperature Gradient (TT > TB)
TTop
TBottom
Figure 4.9: Temperature gradient (TT>TB) case for prestress change analysis
Table 4.5.1: Prestress decrement percentage due to temperature gradient (TT>TB) for 20m
span integral bridge
TB to TT
(c)
30 to 40
20 to 40
10 to 40
0 to 40
Temp.
Gradient
(c)
10
20
30
40
post stress change (kN)

initial prestress
1500
1500
1500
1500
final prestress
1850
2195
2550
2900
prestress
decrement
350
695
1050
1400
prestress
decrement
percentage
23.3
46.3
70.0
93.3
TB to TT
(c)
Temp.
Gradient
(c)
30 to 40
20 to 40
10 to 40
0 to 40
10
20
30
40

initial
prestress
final prestress
1800
2582
1800
3367
1800
4155
1800
4942
prestress
decrement
prestress
decrement
percentage
782
1567
2355
3142
43.4
87.1
130.8
174.6
TB to TT
(c)
30 to 40
20 to 40
10 to 40
0 to 40
Temp.
Gradient
(c)
10
20
30
40

initial
prestress
final prestress
2800
4500
2800
6180
2800
7877
2800
9571
prestress
decrement
1700
3380
5077
6771
prestress
decrement
percentage
60.7
120.7
181.3
241.8
75
Figure 4.10: Graph of prestress decrement percentage versus temperature gradient (TT>TB)
4.4.2 Case 2: Temperature Gradient (TT < TB)
TTop
TBottom
Figure 4.11: Temperature gradient (TT<TB) case for prestress change analysis
76
Table 4.6.1: Prestress decrement percentage due to temperature gradient (TT<TB) for 20m
TB to TT
(c)
Temp.
Gradient
(c)
40-30
40-20
40-10
40-0
10
20
30
40

initial
final
prestress
prestress
1500
1150
1500
805
1500
450
1500
100
prestress
change
prestress
increment
percentage
350
695
1050
1400
23.3
46.3
70.0
93.3
TB to TT
(c)
40-30
40-20
40-10
40-0
Temp.
Gradient
(c)
10
20
30
40

initial
final
prestress
prestress
1800
1032
1800
276
1800
-491
1800
-1260
prestress
change
768
1524
2291
3060
prestress
increment
percentage
42.7
84.7
127.3
170.0
TB to TT
(c)
Temp.
Gradient
(c)
40-30
40-20
40-10
40-0
10
20
30
40

initial
final
prestress
prestress
2800
1134
2800
-500
2800
-2100
2800
-3700
prestress
change
prestress
increment
percentage
1666
3300
4900
6500
59.5
117.9
175.0
232.1
77
Figure 4.12: Graph of prestress increment percentage versus temperature gradient (TT<TB)
4.4.3 Case 3: Uniform Temperature Increment
Figure 4.13: Uniform temperature increment case for prestress change analysis
78
Table 4.7.1: Prestress increment percentage due to uniform temperature increment for 20m
Temp. change
(c)
Temp.
Gradient
(c)
30 to 40
20 to 40
10 to 40
0 to 40
10
20
30
40

initial
final
prestress
prestress
1500
2050
1500
2605
1500
3175
1500
3730
prestress
change
prestress
increment
percentage
550
1105
1675
2230
36.7
73.7
111.7
148.7
Temp. change
(c)
30 to 40
20 to 40
10 to 40
0 to 40
Temp.
Gradient
(c)
10
20
30
40

initial
final
prestress
prestress
1800
2340
1800
2882
1800
3426
1800
3969
prestress
change
540
1082
1626
2169
prestress
increment
percentage
30.0
60.1
90.3
120.5
Temp. change
(c)
Temp.
Gradient
(c)
30 to 40
20 to 40
10 to 40
0 to 40
10
20
30
40

initial
final
prestress
prestress
2800
3462
2800
4121
2800
4788
2800
5445
prestress
change
prestress
increment
percentage
662
1321
1988
2645
23.6
47.2
71.0
94.5
79
Figure 4.14: Graph of prestress decrement percentage versus uniform temperature increment
4.4.4 Case 4: Uniform Temperature Decrement
Figure 4.15: Uniform temperature decrement case for prestress change analysis
80
Table 4.8.1: Prestress decrement percentage due to uniform temperature decrement for 20m
Temp. change
(c)
Temp.
Gradient
(c)
40-30
40-20
40-10
40-0
10
20
30
40

initial
final
prestress
prestress
1500
950
1500
395
1500
-183
1500
-760
prestress
change
prestress
decrement
percentage
-550
-1105
-1683
-2260
-36.7
-73.7
-112.2
-150.7
Temp. change
(c)
40-30
40-20
40-10
40-0
Temp.
Gradient
(c)
10
20
30
40

initial
final
prestress
prestress
1800
1266
1800
745
1800
219
1800
-310
prestress
change
-534
-1055
-1581
-2110
prestress
decrement
percentage
-29.7
-58.6
-87.8
-117.2
Temp. change
(c)
Temp.
Gradient
(c)
40-30
40-20
40-10
40-0
10
20
30
40

initial
final
prestress
prestress
2800
2136
2800
1493
2800
857
2800
230
prestress
change
prestress
decrement
percentage
-664
-1307
-1943
-2570
-23.7
-46.7
-69.4
-91.8
81
Figure 4.16: Graph of prestress decrement percentage versus uniform temperature decrement
82
4.5
Discussion
From the figures, it is shown that the longer the bridge span, the larger and more
obvious the change of prestress in the beam. As the prestress change is linearly analysed from
deflection values, the displacement too shows larger and obvious change of value as the
bridge span is longer.
For temperature gradient (TT>TB) case, the deflection tends to go upwards while for
the temperature gradient (TT<TB) case, the deflection tends to go downwards. Therefore
temperature gradient decrement where the extreme bottom fibre is hotter than the extreme top
fibre, will add to the deflection due to other load cases.
At the same time, for uniform temperature increment case, the deflection tends to go
upwards due to the constraint when it is expanding. Meanwhile, for the uniform gradient
decrement, the deflection tends to go downwards also because it its constrained when
contracting. Therefore, similar to temperature gradient decrement case, uniform temperature
decrement will also add to the deflections due to the other load cases.
From the comparison of deflection subsection, we can clearly see that integral bridge
is better than the simply supported bridge in term of deflection resistant. The deflection for
integral bridge is smaller than the deflection for the simply supported bridge. The uniform
temperature change did not affect deflections for simply supported beam as the beam can
contract and expand freely without constraint.
By varying the temperature and the integral bridge spans in 4 different cases, it is
shown that the prestress value changed. The percentages of these changes vary from 90% to
240%. In Malaysian condition, the temperature usually ranged between 20 C
to 40 C
.
Therefore according to the results, the prestress change percentages in Malaysia will
approximately vary from 46% to 120% considering the extreme conditions.
CHAPTER V
CONCLUSION AND RECOMMENDATION
5.1
Conclusion
The conclusion will be best described by recounting each specific research
objective, research questions and also the accomplishment of the aims:
Objective 1 To study the effect of temperature gradient on the prestressing force in

prestressed integral bridge beam
From the results, it is determined that from the effect of temperature gradient
(TT>TB) the deflection of the beam will be lessened but on the other hand, the
temperature gradient (TT<TB) will cause more deflection to the beam. The
prestressing force is decreased when the beam bends upwards as it is relaxed. On the
opposite, the prestressing force is increased when the beam bends downwards. When
the beam bends downwards, additional stress is added to the tendon.
84
Objective 2 To study the effect of uniform temperature change on the prestressing
force in prestressed integral bridge beam
For uniform temperature change, the beam contracts and expands depending
on whether the uniform temperature increases or decreases. When the beam expands
because of the uniform temperature increment, additional stress will be induced to
the tendon. In contra, the stress in the tendon will be decreased as the beam contracts
because of the uniform temperature decrement. The contractions and expansions
phenomenon happen in x-axis. Therefore for simply supported beam, it does not
affect the displacement values. However, when the beam is constrained like the
integral bridge beam, it will affect the displacement values.
Objective 3 To determine whether temperature effects can be neglected in integral

bridge design considering Malaysian condition
At first, the hypothesis is that at Malaysian condition, the temperature range

and temperature ambient is too small, therefore temperature effects can maybe be
neglected in integral bridge design process. However, after this research is done, the
result shows the opposite.
The values of deflection caused by the temperature effects is relevant to

consider, and proves that it cannot be neglected. If neglected, the deflection caused
by the temperature effects may cause the overall displacement of the integral bridge
to exceed the allowable deflection limit. The span of the integral bridge is an
important factor to the beam deflection due to the temperature effects. In other
words, the longer the span, the more deflection will be caused by the temperature
effect. This is consistent to the formula of thermal expansion *(T)*L and also the
deflection formula = L2/9.6 (1/r).
As a conclusion, the longer the integral bridge span, the more obvious the
change in prestress due to the effect of temperature. Considering extreme conditions,
the prestress change in Malaysia ranged from 46% to 120%. It shows that the
85
temperature effect cannot be neglected and is very relevant and important to consider
in Malaysian condition.
5.2
Recommendations
For the future studies several recommendation are proposed of this finite
element analysis in the effect of temperature on prestressed bridge beam using
LUSAS software. The followings are recommendations relevant to this topic.
1) The study can be extended by developing a 3 dimensional model of integral

bridge where it will simulate as close as possible to the real site condition
2) Instead of comparing and verifying the finite element model with the hand
calculation method, it will be a lot better to compare it to experimental
results.
3) Integral bridge is proved to be more dependable as more and more research is
done on it. Therefore, Malaysia should construct more integral bridge in the
future.
86
REFERENCES
1.
OBrien E. J., Keogh D. L. and Lehane B. M. (1999). Bridge Deck Analysis. London,
E & FN Spon
2.
Newhouse C. D et al. (2008) Modeling Early-Age Bridge Restraint Moments: Creep,

Shrinkage, and Temperature Effects. Journal of Bridge Engineering. Vol. 13, No. 5:
432.
3.
Hambly E.C. (1976). Bridge Deck Behaviour. London, Chapman and Hall
4.
Dhir R.K., Newlands M.D. and McCarthy M.J. (2003). Role of Concrete Bridges in
Sustainable Development. USA, ASCE Press
5.
Chatterjee S. (2003). Modern Steel Bridges. UK, Blackwell Science Ltd
6.
Cope R.J. (1987). Concrete Bridge Engineering. London and New York, Elsevier
Applied Science
7.
Kurth H. (1975). Bridges. London, William Clowes & Sons
8.
Abdullah R. (2009). Finite Element Method Course Note. University Technology of

Malaysia
9.
Connal J. (2003) Integral Abutment Bridges Australian and US Practice. Seminar on

Design & Construction of Integral Bridges. Kuala Lumpur.
10. Babu P.V.M and Bhandari N.M. (2006) A Comparative Study of Integral Bridge
versus Simply Supported Bridge. Advances in Bridge Engineering.

Temperature Effect On Prestressed Integral Bridge Beam

Enviado por

Dados do documento

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Temperature Effect On Prestressed Integral Bridge Beam

Enviado por

Direitos autorais:

Formatos disponíveis

PSZ 19:16 (Pind.

UNIVERSITI TEKNOLOGI MALAYSIA

Authors full name :

MUHAMMAD LUTFI BIN OTHMAN

EFFECT OF TEMPERATURE ON PRESTRESSED INTEGRAL

I declare that this thesis is classified as:

(Contains confidential information under the Official Secret

(Contains restricted information as specified by the

I agree that my thesis to be published as online open access

I acknowledged that Universiti Teknologi Malaysia reserves the right as follows:

DR. REDZUAN BIN ABDULLAH

Date : 23 NOVEMBER 2009

Date : 23 NOVEMBER 2009

I hereby declare that I have read this project report

Name of Supervisor: DR. REDZUAN BIN ABDULLAH

THE EFFECT OF TEMPERATURE ON PRESTRESSED INTEGRAL

MUHAMMAD LUTFI BIN OTHMAN

A report submitted in partial fulfillment of the

Faculty of Civil Engineering

: MUHAMMAD LUTFI BIN OTHMAN

Karya ini adalah dedikasi teristimewa buat emak yang

In this opportunity, I would like to express my gratitude towards my supervisor

Most bridges in Malaysia are pre-stressed concrete type and constructed as

Kebanyakan jambatan di Malaysia ialah jenis konkrit pra-tegasan dan dibina

Scope of The Research

BACKGROUND STUDY OF BRIDGE ANALYSIS

AND INTEGRAL BRIDGE

Integral Bridge Issues

Integral Bridge Limitations

Integral Bridge Construction

Bridge Deck Analysis

Simply Supported Beam/Slab

Series of Simply Supported Beams/Slabs

Continuous Beam/Slab With Full Propping

Partially Continuous Beam/Slab

Frame/Box Culvert (Integral Bridge)

Dead and Superimposed Dead Loading

Imposed Traffic Loading

Imposed Loading Due to Road Traffic

Adaptation of Real Integral Bridge Design

Support Connection Details

LUSAS Structural Modelling

Prestress Definition to BS5400

Prestress Change Determinationby Trial and Error

Comparison of LUSAS and Hand Calculation Trial

and Error Method

Example Calculation for Hand Calculation

RESULT AND DISCUSSION

Case 1: Temperature Gradient (TT>TB)

Case 2: Temperature Gradient (TT<TB)

Case 3: Uniform Temperature Increment

Case 4: Uniform Temperature Decrement

Prestress Change in Prestressed Integral Bridge Beam

Case 1: Temperature Gradient (TT>TB)

Case 2: Temperature Gradient (TT<TB)

Case 3: Uniform Temperature Increment

Case 4: Uniform Temperature Decrement

CONCLUSION AND RECOMMENDATION

Safety factor calculated according to each bridge

General details for beam span 20m, 30m and 40m

Section properties for 20m, 30m and 40m integral

Values of spring stiffness

Trial and error method