PARAMETRIC STUDY OF A CLASS OF ARCH BRIDGE MODELS

By
Ramez Bou Rizk
Ahmad Mokbel
Charbel Bacha

A thesis submitted to the Department of Civil Engineering in partial fulfillment of the

requirements for the degree of Master of Science in Civil Engineering

Faculty of Engineering
University of Balamand

June 2018

Copyright © 2018, Ramez Bou Rizk, Ahmad Mokbel & Charbel Bacha
All Rights Reserved
 iii

ACKNOWLEDGEMENTS

This report could not have been written if it were not for the contribution and
encouragement of various people.
To begin with, much appreciation and gratitude go to Dr. Nabil Fares, the project
supervisor, for his advice, extensive work, and infinite support.
The writers would like to thank the moderators Dr. Najib Saliba and Eng. Philippe
Hawi for their support and advice during the writing of this report.
This report would have never been completed without the encouragement and faith of
families and friends.
 iv

ABSTRACT

This project encompasses a comprehensive parametric study of a class of arch bridge

models in which several numerical parameters are studied in order to understand the
magnitude of their effect when varied. The method of choice to analyze the various structures
is the Direct Stiffness Method, and this is implemented in both manual and automated
calculations. The aim of this project is to optimize the parameters related to stress distribution
and select the best possible configuration for a class of arch bridge models. The major
constraints studied include: overall dimensions, member stiffness factors, support-ends,
applied forces, deflections, and bending moments.
The project begins with basic idealized arch bridge examples solved manually, and
these serve as control examples that aid in later stages for verification. The next step includes
developing a program to facilitate the analysis since several iterations are expected in order to
reach an optimal result. The program is developed using the Wolfram Mathematica platform
and includes an automated algorithm to assist in analyzing several cases so that a conclusive
result can be achieved. The application of a parametric study easily allows the identification
of critical variable. This is important since the final outcome is to specify certain parameters
and obtain maximal efficiency and capability of a half-through arch bridge.
 v

TABLE OF CONTENTS

ACKNOWLEDGEMENTS iii
ABSTRACT iv
TABLE OF CONTENTS v
LIST OF ABBREVIATIONS vii
LIST OF TABLES viii
LIST OF FIGURES ix
CHAPTER 1: INTRODUCTION 1
1.1 General Introduction 1
1.2 Background on Arch Bridges 1
1.2.1 Applications of Arch Bridges 1
1.2.2 Benefits of Arch Bridges 2
1.2.3 Types of Arch Bridges 3
1.3 Aims and Objectives 5
1.4 Methodology Chart 6
1.5 Scope of Work 7
CHAPTER 2: MODELING OF ARCH BRIDGES 8
2.1 Basic Approach to Solve Simple Cases 8
2.2 Simple Cases Considered 8
2.3 Simple Case Solutions with Discussions 9
2.3.1 Case 1a: Two-Member Truss over Beam (Pinned Supports) 9
2.3.2 Case 1b: Two-Member Truss over Beam (Pinned and Roller Supports) 13
2.3.3 Case 2a: Two-Member Frame over Beam (Pinned Supports) 18
CHAPTER 3: COMPUTER METHOD FOR SOLVING GENERAL CASE 23
3.1 Direct Stiffness Method 23
3.2 Verification of Simple Cases with SAP 24
3.2.1 Modeling with SAP 24
3.2.2 Verification of Case 1a 31
3.2.3 Verification of Case 1b 34
3.2.4 Verification of Case 2a 37
 vi

3.3 Arc of Circle Arch 40

3.3.1 Definition of Arc of Circle Arch 40
3.3.2 Limitations of Arc of Circle Arch 41
CHAPTER 4: PARAMETRIC STUDIES OF ARC OF CIRCLES 42
4.1 Range of Values Considered 42
4.2 Results and Discussions 45
4.2.1 Study 1: Pinned Arch and Deck - Hinged Connection - EI Arch = 0.1 45
4.2.2 Study 2: Pinned Arch and Deck - Hinged Connection - EI Arch = 1 48
4.2.3 Study 3: Pinned Arch and Deck - Hinged Connection - EI Arch = 10 51
4.2.4 Study 4: Pinned Arch and Deck - Stiff Connection - EI Arch = 0.1 54
4.2.5 Study 5: Pinned Arch and Deck - Stiff Connection - EI Arch = 1 57
4.2.6 Study 6: Pinned Arch and Deck - Stiff Connection - EI Arch = 10 60
4.3 Discussions of Parametric Study 63
CHAPTER 5: CONCLUSIONS AND RECOMMENDATIONS 65
5.1 Conclusions 65
5.2 Recommendations 65
LIST OF REFERENCES 66
 vii

LIST OF ABBREVIATIONS

A Arch Arch Cross-Sectional Area

A Beam (Deck) Beam (Deck) Cross-Sectional Area
A Truss Truss (Inclined Member) Cross-Sectional Area
C Cosine θ
CCW Counter Clockwise
DOF Degree of Freedom
E Young’s Modulus (Modulus of Elasticity)
F Axial Force
H Top Height from Beam (Deck) to Arch Top
I Arch Arch Moment of Inertia
I Beam (Deck) Beam (Deck) Moment of Inertia
I Truss Truss (Inclined Member) Moment of Inertia
L Total Bridge Span
L Mid Bridge Span between Points of Arch Intersection
M Bending Moment
P Applied Point Load
r Radius of Gyration
S Sin θ
SAP Structural Analysis Program
UX Horizontal Degree of Freedom
UY Vertical Degree of Freedom
v Modules of Shear (Modulus of Rigidity)
αt Relative Axial Stiffness of Truss to Beam (Deck) (EAt / EAb)
β Relative Truss to Beam (Deck) Stiffness (EAt × L2 / EIb)
δ Mid Deflection at Middle of Beam (Deck)
Θ Rotational Degree of Freedom
θ Angle between Arch and Beam (Deck) Measured CCW from Horizontal
 viii

LIST OF TABLES

Table 2.1 LM Vectors of Two-Member Truss over Beam (Pinned Supports). 10

Table 2.2 LM Vectors of Two-Member Truss over Beam (Pinned and Roller 15
Supports).
Table 2.3 LM Vectors of Two-Member Frame over Beam (Pinned Supports). 19
Table 3.1 Numeric Values Assigned for SAP Modeling. 25
Table 3.2 Case 1a Difference between Derived Formulas and SAP Results. 34
Table 3.3 Case 1b Difference between Derived Formulas and SAP Results. 37
Table 3.4 Case 2a Difference between Derived Formulas and SAP Results. 40
Table 4.1 Numeric Values Assigned for Mathematica Model. 43
Table 4.2 Pinned Arch and Deck – Various Arch Layouts. 44
Table 4.3 Pinned Arch and Deck – Various Arch Layouts with Deformation. 44
Table 4.4 Pinned Arch and Deck - Hinged - EI Arch = 0.1 - Arch Moment. 45
Table 4.5 Pinned Arch and Deck - Hinged - EI Arch = 0.1 - Deck Moment. 46
Table 4.6 Pinned Arch and Deck - Hinged - EI Arch = 0.1 - Deflection. 47
Table 4.7 Pinned Arch and Deck - Hinged - EI Arch = 1 - Arch Moment. 48
Table 4.8 Pinned Arch and Deck - Hinged - EI Arch = 1 - Deck Moment. 49
Table 4.9 Pinned Arch and Deck - Hinged - EI Arch = 1 - Deflection. 50
Table 4.10 Pinned Arch and Deck - Hinged - EI Arch = 10 - Arch Moment. 51
Table 4.11 Pinned Arch and Deck - Hinged - EI Arch = 10 - Deck Moment. 52
Table 4.12 Pinned Arch and Deck - Hinged - EI Arch = 10 - Deflection. 53
Table 4.13 Pinned Arch and Deck - Stiff - EI Arch = 0.1 - Arch Moment. 54
Table 4.14 Pinned Arch and Deck - Stiff - EI Arch = 0.1 - Deck Moment. 55
Table 4.15 Pinned Arch and Deck - Stiff - EI Arch = 0.1 - Deflection. 56
Table 4.16 Pinned Arch and Deck - Stiff - EI Arch = 1 - Arch Moment. 57
Table 4.17 Pinned Arch and Deck - Stiff - EI Arch = 1 - Deck Moment. 58
Table 4.18 Pinned Arch and Deck - Stiff - EI Arch = 1 - Deflection. 59
Table 4.19 Pinned Arch and Deck - Stiff - EI Arch = 10 - Arch Moment. 60
Table 4.20 Pinned Arch and Deck - Stiff - EI Arch = 10 - Deck Moment. 61
Table 4.21 Pinned Arch and Deck - Stiff - EI Arch = 10 - Deflection. 62
Table 4.22 Various Parametric Study Coefficients from Study 1. 63
 ix

LIST OF FIGURES

Figure 1.1 Sydney Harbour Bridge (Half-Through Arch Bridge). 2

Figure 1.2 Bayonne Bridge (Half-Through Arch Bridge). 3
Figure 1.3 Types of Arch Bridges. 4
Figure 1.4 Methodology Chart. 6
Figure 2.1 Simple Cases Considered. 8
Figure 2.2 Two-Member Truss over Beam (Pinned Supports). 9
Figure 2.3 Angle at Which Either the Force in the Truss or Bending Moment in 11
the Beam is a Maximum Verses the Relative Truss to Beams
Stiffness.
Figure 2.4 Maximum Force in Truss and Maximum Bending Moment in Beam 12
at Optimal Angle as a Function of the Relative Truss to Beam
Stiffness.
Figure 2.5 Case 1a Normalized Deflection at Beam Middle for Various Angles 13
as β is Varied.
Figure 2.6 Two-Member Truss over Beam (Pinned and Roller Supports). 14
Figure 2.7 Change in Optimal Angle at which the Force in the Truss is a 16
Maximum Verses the Relative Truss to Beams Stiffness.
Figure 2.8 Change in Optimal Angle at Which the Bending Moment in the Beam 17
is a Maximum Verses the Relative Truss to Beams Stiffness.
Figure 2.9 Case 1b Normalized Deflection at Beam Middle for Various Angles 18
as β is Varied.
Figure 2.10 Two-Member Frame over Beam (Pinned Supports). 19
Figure 2.11 Case 2a Normalized Deflection at Beam Middle for Various Angles 20
as β is Varied.
Figure 2.12 Case 2a Normalized Axial Force in Inclined Frame for Various 21
Angles as β is Varied.
Figure 2.13 Case 2a Normalized Bending Moment at Beam Middle for Various 22
Angles as β is Varied.
Figure 3.1 Truss Member Stiffness Matrix. 23
Figure 3.2 Beam with No Axial Force Stiffness Matrix. 23
Figure 3.3 General Frame Member Stiffness Matrix. 24
 x

Figure 3.4 Beam Section Dimensions for SAP Modeling. 25

Figure 3.5 Truss Section Dimensions for SAP Modeling. 26
Figure 3.6 SAP Version and Properties Window. 26
Figure 3.7 Model Unit and Grid System Initialization. 27
Figure 3.8 Grid System Dimensions. 27
Figure 3.9 Beam and Truss Material Properties. 28
Figure 3.10 Adjusting Self Weight Multiplier to Neglect Element Weights. 28
Figure 3.11 Beam Section Dimensions Based on Reduced Radius of Gyration. 29
Figure 3.12 Beam Section Properties Based on Selected Dimensions. 29
Figure 3.13 Truss Section Dimensions Based on Reduced Radius of Gyration. 30
Figure 3.14 Truss Section Properties Based on Selected Dimensions. 30
Figure 3.15 Infinitely Stiff Link between Arch Top and Beam Middle. 31
Figure 3.16 Case 1a General Layout and Applied Load. 32
Figure 3.17 Case 1a Member and Hinge Layout. 32
Figure 3.18 Case 1a Vertical Deflection. 33
Figure 3.19 Case 1a Bending Moment Diagram. 33
Figure 3.20 Case 1a Axial Force Diagram. 34
Figure 3.21 Case 1b General Layout and Applied Load. 35
Figure 3.22 Case 1b Member and Hinge Layout. 35
Figure 3.23 Case 1b Vertical Deflection. 36
Figure 3.24 Case 1b Bending Moment Diagram. 36
Figure 3.25 Case 1b Axial Force Diagram. 37
Figure 3.26 Case 2a Member and Hinge Layout. 38
Figure 3.27 Case 2a Vertical Deflection. 38
Figure 3.28 Case 2a Bending Moment Diagram. 39
Figure 3.29 Case 2a Axial Force Diagram. 39
Figure 3.30 Half-Through Arch Bridge Dimensions. 41
Figure 4.1 Pinned Arch and Deck - Hinged - EI Arch = 0.1 - Arch Moment. 45
Figure 4.2 Pinned Arch and Deck - Hinged - EI Arch = 0.1 - Deck Moment. 46
Figure 4.3 Pinned Arch and Deck - Hinged - EI Arch = 0.1 - Deflection. 47
Figure 4.4 Pinned Arch and Deck - Hinged - EI Arch = 1 - Arch Moment. 48
Figure 4.5 Pinned Arch and Deck - Hinged - EI Arch = 1 - Deck Moment. 49
 xi

Figure 4.6 Pinned Arch and Deck - Hinged - EI Arch = 1 - Deflection. 50

Figure 4.7 Pinned Arch and Deck - Hinged - EI Arch = 10 - Arch Moment. 51
Figure 4.8 Pinned Arch and Deck - Hinged - EI Arch = 10 - Deck Moment. 52
Figure 4.9 Pinned Arch and Deck - Hinged - EI Arch = 10 - Deflection. 53
Figure 4.10 Pinned Arch and Deck - Stiff - EI Arch = 0.1 - Arch Moment. 54
Figure 4.11 Pinned Arch and Deck - Stiff - EI Arch = 0.1 - Deck Moment. 55
Figure 4.12 Pinned Arch and Deck - Stiff - EI Arch = 0.1 - Deflection. 56
Figure 4.13 Pinned Arch and Deck - Stiff - EI Arch = 1 - Arch Moment. 57
Figure 4.14 Pinned Arch and Deck - Stiff - EI Arch = 1 - Deck Moment. 58
Figure 4.15 Pinned Arch and Deck - Stiff - EI Arch = 1 - Deflection. 59
Figure 4.16 Pinned Arch and Deck - Stiff - EI Arch = 10 - Arch Moment. 60
Figure 4.17 Pinned Arch and Deck - Stiff - EI Arch = 10 - Deck Moment. 61
Figure 4.18 Pinned Arch and Deck - Stiff - EI Arch = 10 - Deflection. 62
 1

CHAPTER 1
INTRODUCTION

1.1 General Introduction

One of the major fields of civil engineering is the transportation field. This includes

roads, railroads, and other facilities which all rely on bridges when the geography of the

location includes steep slopes and rivers. The idea of bridges evolved from the notion of a

tree-trunk over a stream to a vast array of complex structures that meet different

requirements. The Romans refined bridge building by implementing arch bridges because

they discovered that arches can support more weight than a flat structure, and this innovation

led to the evolution of arch bridges.

Arch bridges are one of the oldest types of bridges that have been used for a long time.

This is due to their unique mechanism of transferring loads to the supports which have

resulted in great natural strength. Before the introduction of technology to the civil

engineering field, arch bridges were built in a non-optimal way; which means that all of the

elements were not being used to their maximum capability. Today with the help of

technology, the design of different structures is being done with the use of numerous

programs which have facilitated the work to help assure that structures built are executed the

right and proper way. This project aims to analyze and optimize a class of half-through arch

bridge.

1.2 Background on Arch Bridges

1.2.1 Applications of Arch Bridges

Arch bridges were first built from stone, bricks, clay, and timber. These materials were

later replaced by concrete and steel. Theories regarding the strength of building materials
 2

were developed during the Renaissance by Leonardo da Vinci and Galileo Galilei, who were

both polymaths. Advancements in the educational field lead to new majors in fields such as

engineering, scientists, and mathematicians. These specialists worked out the mechanics of

bridges. Our point of interest is a class of half-through arch bridges since it will be seen that

there are several categories of arch bridges. Arch bridges are particularity unique because

they draw their strength from the compression of the arch and the lateral confinement of the

end supports.

Figure 1.1: Sydney Harbour Bridge (Half-Through Arch Bridge) [5].

1.2.2 Benefits of Arch Bridges

The major benefits of arch bridges include:

 Provide High Level of Resistance – The curved design of the arch gives an

extraordinary point of strength. It allows a higher level of resistance to the bending

forces that may try to deform it. Due to the symmetry of the structure, the horizontal

forces that are placed on it are displaced equally [7].

 Resistant to Distortion – An arch bridge with an arc of circle arch is purposely designed

to ensure that no damage or distortion would occur to the bridge under excessive
 3

amounts of weight. This is very beneficial since it helps cut down potential maintenance

cost [7].

 Uniform Pressure Distribution – The weight of anything crossing an arch bridge will be

transferred from the arch to the end supports, instead of dispersing across their entirety.

This is done to make sure no single part of the bridges takes on too much pressure [7].

 Gain Strength Over Time – Since arch bridges are built to resist compression the bridge

would actually become stronger over time. This is a material related phenomenon since

the atoms of a material under constant compression move closer and gain compressive

strength [7].

Figure 1.2: Bayonne Bridge (Half-Through Arch Bridge) [4].

1.2.3 Types of Arch Bridges

In arch bridges, the forces are transmitted along from the top member all the way to the

supports located at each end, unlike straight bridges where the force is concentrated at a

certain point. The critical component to transfer forces between all the connections is the
 4

hinges at the intersections of the arch and deck which do not allow transferring of the

moment. There are three types of arch bridges, illustrated in Figure 1.3, and these include:

 Deck Arch Bridge

 Through Arch Bridge

 Half-Through Arch Bridge

Figure 1.3: Types of Arch Bridges [8].

 Deck Arch Bridge (a) – The spandrel columns are the main support of the arch ribs. It is

usually made of reinforced concrete and the deck is located above the crown of the

arch.
 5

 Through Arch Bridge (b) – The suspenders are the support and the deck is located

below the arch for the entire span.

 Half-Through Arch Bridge (c) – The spandrel columns and suspender are the supports

of bridge deck and it is used for long spans.

1.3 Aims and Objectives

The aim of this project is to optimize parameters related to stress distribution and select

the best possible configuration for a class of arch bridge models. The objectives of this

project are to:

 Solve basic arch bridge models using manual calculations that will be used as

references.

 Develop an automated algorithm using the Wolfram Mathematica platform capable of

calculating critical parameters for several cases in a bid to optimize the overall

structure.
 6

1.4 Methodology Chart

Figure 1.4: Methodology Chart.

 7

1.5 Scope of Work

The basic scope of work that will be accomplished includes:

 Solving basic examples that consist of a simple truss over a beam with various end

support conditions.

 Calculating deflections, moments, and forces for the basic examples and creating plots

that aim to correlate important parameters.

 Verify manual calculations by modeling the basic examples on SAP.

 A program is written using Wolfram Mathematica to study half-through arch bridges.

 Identify parameter limitations and conduct a parametric study. All results are tabulated

as functions of normalized variables.

 Propose potential recommendations and possible improvements to further advance this

research.
 8

CHAPTER 2
MODELING OF ARCH BRIDGES

2.1 Basic Approach to Solve Simple Cases

In this project, a rudimentary approach is used to analyze basic forms of arch bridges.

This includes understanding the various elements (members, supports, and connections) as

well as their different properties (stiffness, dimensions, and orientation of members) that

make up a typical arch bridge. These simple cases are idealized forms of arch bridges that can

be used as reference models when studying more complicated examples and as controlling

cases when testing an automated program in later stages of the project. In each example, a

figure is displayed to aid in understanding the case and followed by various plots that are

further explained.

2.2 Simple Cases Considered

A total of three simple cases are considered in this project. These cases are each made

up of three members and have two supports on either end. The simplest possible form of an

arch is used, and this involves two members in a triangular configuration. A single point load

of magnitude ‘P’ is placed at the center of the beam. In order to ensure load distribution

between the arch and beam, an infinitely stiff link is placed connecting them together. Figure

2.1 illustrates the three basic examples that will be covered in more details in the next section.

Figure 2.1: Simple Cases Considered.

 9

2.3 Simple Case Solutions with Discussions

2.3.1 Case 1a: Two-Member Truss over Beam (Pinned Supports)

The first simple case solved consists of a two-member truss (AD and CD) over a beam

(AB and BC) as seen in the figure below. The case will be solved using manual application of

the direct stiffness method and the final results will include the deflection and moments at

mid-span in the beam and the axial load in the truss elements.

The structure has a single point load applied downwards at the center of the beam with

a magnitude of ‘P’. An infinite stiff element is placed between nodes B and D to transfer the

loads between the truss elements and beam. The supports at A and C are pinned, therefore

restricting lateral and vertical translation while allowing rotation.

Figure 2.2: Two-Member Truss over Beam (Pinned Supports).

 10

This fundamental arch bridge has a total of three degrees of freedom. It initially starts

with four degrees of freedom, but due to the infinite stiff element between nodes B and D the

vertical degrees of freedom are equal at these nodes. Additionally, each support has a

rotational degree of freedom.

The direct stiffness method is used to analyze the above case and the first step is to

construct the LM vectors of the elements. Axial deformations are ignored in members AB

and BC since they are laterally restricted and this is shown in the table below.

Table 2.1: LM Vectors of Two-Member Truss over Beam (Pinned Supports).

LM UXS UYS ΘS UXE UYE ΘE

AB - 0 2 - 1 0
BC - 1 0 - 0 3
AD 0 0 - 0 1 -
CD 0 0 - 0 1 -

The following simplified results are obtained:

1 PL3
δMid = 1 EA L2
× (2.1)
1+12× EIt Cos[θ]Sin[θ]2 48 EIb
b

1 PL
𝑀Mid = 1 EA L2
× (2.2)
1+12× EIt Cos[θ]Sin[θ]2 4
b

1 EAt L2 Sin[𝜃]Cos[𝜃] P
𝐹Truss = 1 EA L2
× × (2.3)
1+12× EIt Cos[θ]Sin[θ]2 12EIb 2
b

We note that the dependence of δMid and 𝑀Mid on θ are exactly the same. Due to the

symmetry of the structure, the maximum moment and deflection occur at the middle of the
 11

beam. Using the derived equations, we were able to create three plots that study the various

parameters as the angle and relative truss to beam stiffness (β) is varied.

Figure 2.3 shows the angle required to obtain the maximum force and bending moments

as the relative truss to beam stiffness (β) is varied. It is noticed that the optimal angle to

obtain the maximum bending moment in the beam is constant and independent of the relative

beam to truss stiffness ratio. To reach the maximum force in the truss as β is increasing, a

smaller angle is required. This is because as EAt is increasing the truss is capable of resisting

more load. Finally, we note that at 45 degrees the force in the truss is maximum for a very

stiff beam resulting in a β value of zero.

Figure 2.3: Angle at Which Either the Force in the Truss or Bending Moment in the Beam is a Maximum

Verses the Relative Truss to Beams Stiffness.

 12

There is a rapid change in angle when β increases from zero to 200. This is due to the

increase in the axial stiffness of the truss leading it to resist a maximum force at a shallower

angle. In conclusion, to obtain the maximum force in the truss a lower angle is required as the

relative stiffness ratio increases.

Figure 2.4 illustrates the distribution of forces among the members while varying the

relative truss to beam stiffness (β). When β has a value of zero and due to the truss having no

axial stiffness, the beam has a maximum bending moment value. Conversely, when the β is

large the truss has a maximal value while the beam is minimal. As the relative stiffness ratio

approaches infinity, the truss tends to resists all the applied load.

Figure 2.4: Maximum Force in Truss and Maximum Bending Moment in Beam at Optimal Angle as a

Function of the Relative Truss to Beam Stiffness.

 13

Using Figure 2.5 we notice that at zero degrees, the deflection is maximum because the

truss has no contribution, and the system is acting as if it is a simply supported beam. It can

also be seen that at 90 degrees a similar behavior is noted, but in this case, the truss has a

length that tends to infinity resulting in a complete loss of stiffness. The optimal angle is seen

to be between 45 to 60 degrees. In this case, the beam and truss are both carrying the load in a

manner that the deflection is minimized. As β increases, the deflection decreases since the

truss is able to support larger loads.

Figure 2.5: Case 1a Normalized Deflection at Beam Middle for Various Angles as β is Varied.

2.3.2 Case 1b: Two-Member Truss over Beam (Pinned and Roller Supports)

The second simple case solved consists of a two-member truss (AD and CD) over a

beam (AB and BC) as seen in the figure below. The same method of solving and results will

be obtained as in the first example.

 14

The structure has a single point load applied downwards at the center of the beam with

a magnitude of ‘P’. An infinite stiff element is placed between nodes B and D to transfer the

loads between the truss elements and beam. The support at A is pinned while C is a roller.

They both restrict vertical translation and allow rotation. Additionally, support C allows

lateral translation, unlike support A. This case has a total of five degrees of freedom. The first

three degrees of freedom are similar to the previous case, in addition to two horizontal

degrees of freedom at nodes B and C.

Figure 2.6: Two-Member Truss over Beam (Pinned and Roller Supports).

The direct stiffness method is also used to analyze the above case and the first step is to

construct the LM vectors of the elements. In this case, axial deformations are not ignored in
 15

members AB and BC, since laterally deformation is permitted by the roller at node C. This is

shown in the LM vectors below.

Table 2.2: LM Vectors of Two-Member Truss over Beam (Pinned and Roller Supports).

LM UXS UYS ΘS UXE UYE ΘE

AB 0 0 4 2 1 0
BC 2 1 0 3 0 5
AD 0 0 - 0 1 -
CD 3 0 - 0 1 -

The following simplified results are obtained:

1 𝑃𝐿3
δMid = 1 EAt𝐿2 (1+αtCos[𝜃]3 )
× (2.4)
1+12× Cos[𝜃]Sin[𝜃]2 48 EIb
EIb (1+2αtCos[𝜃]3 )

1 PL
𝑀Mid = 1 EA 𝐿2 (1+αtCos[𝜃]3 )
× (2.5)
1+ × t Cos[𝜃]Sin[𝜃]2 4
12 EIb (1+2αtCos[𝜃]3 )

EAt 𝐿2
EIb 1 P
𝐹Truss = Cos[𝜃]Sin[𝜃] 1 EAt𝐿2 (1+αtCos[𝜃]3 )
× (2.6)
12 1+ × Cos[𝜃]Sin[𝜃]2 2
12 EIb (1+2αtCos[𝜃]3 )

A similar dependence between δMid and 𝑀Mid on θ is noted as in the previous case.

Additionally, due to the symmetry of the structure, maximum moment and deflection occur at

the middle of the beam. Using the obtained results, we were able to create three plots that

study the various parameters as the angle and relative truss to beam stiffness (β) is varied.

Figure 2.7 shows the change in optimal angle needed to obtain the maximum force in

the truss. The relative truss to beam stiffness is varied along with various values of relative

axial stiffness of truss to beam (αt). As EAb tends to infinity, αt approaches zero and it is seen
 16

that the change in optimal angle is zero. This is because the beam has no axial deformation

and reverts back to the previous example. For αt ranging between 0.01 to 0.2, which are

practical values, the maximum change in angle ranges from 0.85 to -0.3 degrees which is not

a major difference from the previous case. Therefore, we can conclude that the effect of the

roller support is not critical and it would be easier for practical purposes to ignore the axial

deformation in the beam and hence considered the supports of a bridge strictly as pins.

Figure 2.7: Change in Optimal Angle at which the Force in the Truss is a Maximum Verses the Relative

Truss to Beams Stiffness.

Similarly, Figure 2.8 shows the change in optimal angle needed to obtain the maximum

bending moment at the middle of the beam. A similar conclusion can be drawn since the

change in angle varies from 1.3 to 0.9 degrees as αt is varied. This is also not a major

difference from the previous case.

 17

Figure 2.8: Change in Optimal Angle at Which the Bending Moment in the Beam is a Maximum Verses

the Relative Truss to Beams Stiffness.

Similar to case 1a we notice for example 1b, in Figure 2.9, that at zero degrees, the

deflection is maximum because the truss is not helping at all and the structure is acting as if it

is a simply supported beam. It can also be seen that at 90 degrees a similar behavior is noted,

but in this case, the truss has a length that tends to infinity resulting in a complete loss of

stiffness.

The optimal angle is seen to be between 45 to 60 degrees. In this case, the beam and

truss are both carrying the load in a manner that the deflection is minimized. As β increases,

the deflection decreases due to a higher ratio of the truss to beam ratio since the truss begins

to resist larger loads.

 18

Figure 2.9: Case 1b Normalized Deflection at Beam Middle for Various Angles as β is Varied.

2.3.3 Case 2a: Two-Member Frame over Beam (Pinned Supports)

The final simple case solved consists of a two-member frame (AD and CD) over a beam

(AB and BC) as seen in the figure below. The same method of solving and results will be

obtained as in the previous two examples.

The loading and element layout are similar to the previous cases, while the supports at

A and C are pinned. Therefore, restricting lateral and vertical translation while allowing

rotation. This case has a total of three degrees of freedom that are similar to those of the first

example presented in this chapter. It is worth noting that since the inclined elements are

frames and not trusses in this example an additional rotational degree of freedom is present at

node D. This rotational DOF is not shown since it is zero due to the symmetry of the structure

and loading.
 19

Figure 2.10: Two-Member Frame over Beam (Pinned Supports).

The direct stiffness method is used to analyze the above case and the first step is to

construct the LM vectors of the elements. Axial deformations are ignored in members AB

and BC, since they are laterally restricted and this is shown in the LM vectors below.

Table 2.3: LM Vectors of Two-Member Frame over Beam (Pinned Supports).

LM UXS UYS ΘS UXE UYE ΘE

AB - 0 2 - 1 0
BC - 1 0 - 0 3
AD 0 0 2 0 1 0
CD 0 0 3 0 1 0
 20

The following simplified results are obtained:

1 PL3
δMid = 1 EAt𝐿2
× (2.7)
3
(10+22Cos[𝜃]−7Cos[3𝜃]+6Cos[4𝜃]+Cos[5𝜃])+ × Cos[𝜃]Sin[𝜃]2 48EIb
48 12 EIb

(−1−Cos[𝜃]+Cos[𝜃]2 ) PL
𝑀Mid = 1 × (2.8)
(1+3Cos[𝜃]−3Cos[𝜃]2 −3Cos[𝜃]3 +3Cos[𝜃]4 +Cos[𝜃]5 )+12𝛽Cos[𝜃]Sin[𝜃]2 4

2𝛽Sin[2𝜃] P
𝐹Truss = (66+𝛽)Cos[𝜃]−(21+𝛽)Cos[3𝜃]+3(10+6Cos[4𝜃]+Cos[5𝜃]) × (2.9)
2

We notice that at zero degrees in Figure 2.11 the deflection is equal to half, this is due

to the help of the inclined frames that are horizontal in this case. They are adding to the

section of the beam thus doubling flexural resistance and reducing deflection. At 90 degrees,

the frame is not carrying any of the loads thus the beam is resisting all the loads resulting to a

deflection value similar to that of a simply supported beam. This is due to the inclined frame

having a length that tends to infinity resulting in a complete loss of stiffness.

Figure 2.11: Case 2a Normalized Deflection at Beam Middle for Various Angles as β is Varied.
 21

Figure 2.12 illustrates the axial force in the inclined frame members for various values

of β. At an angle of zero, the axial force is zero due to the inclined frames being

perpendicular to the load and hence no axial force is being exerted. A similar resulted is

noted at 90 degrees, but in this case, the inclined frames have no stiffness and cannot resist

any load. The optimal angle occurs near 45 degrees for β up to 10. When the value of β

increases above 100 then the force in the inclined frame is significantly large and this is due

to the increase in their axial stiffness allowing them to resist much larger forces.

Figure 2.12: Case 2a Normalized Axial Force in Inclined Frame for Various Angles as β is Varied.

A similar explanation is valid for Figure 2.13 as that used for Figure 2.11. We also

notice a similarity in the shape of the curves and similar values at zero and 90 degrees. This is

due to the fact that the bending moment and deflection in the beam are directly related.

Hence, at a maximum deflection, we also get the maximum moment.

 22

Regarding the optimal angle, it is found to be in the range of 45 degrees for a β value of

100. While for smaller β values a maximum moment is achieved at 45 degrees. This is the

opposite of Figure 2.12 since the applied load is shared between the inclined frames and

beam.

Figure 2.13: Case 2a Normalized Bending Moment at Beam Middle for Various Angles as β is Varied.
 23

CHAPTER 3
COMPUTER METHOD FOR SOLVING GENERAL CASE

3.1 Direct Stiffness Method

A stiffness method is a structural analysis method where displacements and rotations

are considered to be the primary unknowns. One such example of this is the direct stiffness

method which is a matrix based structural analysis theory. This is particularly suited for

computer-automated analysis of complex structures. It is therefore particularly desirable for

this project since one of our aims is to create an automated program. The results of the direct

stiffness method include: deflections, bending moments and axial forces.

The method works by identifying the degrees of freedom of the elements (LM vector of

each element) and then assembling the stiffness matrix of the structure. The individual

element stiffness matrices are obtained by using Figures 3.1, 3.2, and 3.3.

Figure 3.1: Truss Member Stiffness Matrix [3].

Figure 3.2: Beam with No Axial Force Stiffness Matrix [3].

 24

Figure 3.3: General Frame Member Stiffness Matrix [3].

This method was used to obtain all the equations presented in the previous chapter of

this report. This chapter will verify the credibility of these equations by assigning numeric

values for all parameters and comparing the obtained values with equivalent SAP models of

the analyzed structures.

3.2 Verification of Simple Cases with SAP

3.2.1 Modeling with SAP

Structural analysis program, commonly known as SAP, is a structural modeling

program capable of analyzing complex structures in real time. We modeled the three basic

arch bridge cases defined in the previous chapter of this report. The first step was to assign

numeric values to all the parameters required to calculate deflections, bending moments, and

axial forces. These parameters are summarized in Table 3.1. The results obtained from SAP

are compared to those obtained from the derived direct stiffness equations, and the

differences are noted in order to study the accuracy of the direct stiffness method.
 25

Table 3.1: Numeric Values Assigned for SAP Modeling.

Parameter Numeric Value

A Beam 1.227 m2
A Truss 0.09 m2
E 200 GPa
H 3m
I Beam 0.0171 m4
L 5m
P 100 kN
αt 0.07335
β 1792.54
θ 50.19o

It is worth noting that SAP calculates both flexural and shear deformation stresses. The

direct stiffness method neglects the effect of shear deformation. This issue is solved by using

sections with minimal values of radius of gyration and materials with a large shear modulus.

Figure 3.4: Beam Section Dimensions for SAP Modeling.

 26

Figure 3.5: Truss Section Dimensions for SAP Modeling.

The version of SAP used is 19.2.0 and the following series of figures illustrate the basic

set-up process of the grid system, material, and required sections. These steps are similar for

all the cases, and therefore will only be shown once.

Figure 3.6: SAP Version and Properties Window.

 27

Figure 3.7: Model Unit and Grid System Initialization.

Figure 3.8: Grid System Dimensions.

 28

Figure 3.9: Beam and Truss Material Properties.

Figure 3.10: Adjusting Self Weight Multiplier to Neglect Element Weights.

 29

Figure 3.11: Beam Section Dimensions Based on Reduced r.

Figure 3.12: Beam Section Properties Based on Selected Dimensions.

 30

Figure 3.13: Truss Section Dimensions Based on Reduced r.

Figure 3.14: Truss Section Properties Based on Selected Dimensions.

 31

Figure 3.15: Infinitely Stiff Link between Arch Top and Mid-Beam.

3.2.2 Verification of Case 1a

The first case, as previously defined, involves a two-member truss over a beam with

pinned connections on both ends. This is the simplest of the three cases, and applying the

numeric values defined in Table 3.1 in Equations 2.1, 2.2, and 2.3 we obtained the following

results.

1 100 × 53
δMid = × = 1.481 × 10−5 m
0.09 ×52 48 × 200 × 106 × 0.0171
1 + 12 × 0.0171 Cos[50.19]Sin[50.19]2
 32

1 100 × 5
MMid = × = 24.31 kN. m
0.09 × 52 4
1 + 12 × 0.0171 Cos[50.19]Sin[50.19]2

1 0.09 × 52 × Sin[50.19]Cos[50.19] 100

FTruss = × × = 52.43 kN
0.09 × 52 12 × 0.0171 2
1+ Cos[50.19]Sin[50.19]2
12 × 0.0171

Figure 3.16: Case 1a General Layout and Applied Load.

Figure 3.17: Case 1a Member and Hinge Layout.

 33

Figure 3.18: Case 1a Vertical Deflection.

Figure 3.18 shows the vertical deflection at the highlighted nodes. We note that the

deflections are equal. This is expected since the infinitely stiff link causes the direct transfer

of the applied load and therefore the deflection at the top of the arch equals that of the beam.

Figure 3.19: Case 1a Bending Moment Diagram.

 34

Figure 3.20: Case 1a Axial Force Diagram.

Using Figure 3.19 and 3.20 we note that the bending moment and axial force values are

23.96 kN.m and 52.61 kN respectively.

Table 3.2: Case 1a Difference between Derived Formulas and SAP Results.

Studied Parameters Formula Results SAP Results % Difference

δMid 0.01481 mm 0.01468 mm 0.9 %
MMid 24.31 kN.m 23.96 kN.m 1.5 %
FTruss 52.43 kN 52.61 kN 0.3 %

3.2.3 Verification of Case 1b

The second case involves a two-member truss over beam with a pinned connection at

one end and roller at the other end. The same numeric values defined in Table 3.1 are applied

for Equations 2.4, 2.5 and 2.6. We obtained the following results.

1 100 × 53
δMid = ×
0.09 × 52 (1 + 0.07335 × Cos[50.19]3 ) 48 × 200 × 106 × 0.0171
1 + 12 × 0.0171 Cos[50.19]Sin[50.19]2
(1 + 2 × 0.07335 × Cos[50.19]3 )

δMid = 1.503 × 10−5 m

 35

1 100 × 5
𝑀Mid = 3) ×
0.09 × 52 (1 + 0.07335 × Cos[50.19] 4
1+ Cos[50.19]Sin[50.19]2
12 × 0.0171 (1 + 2 × 0.07335 × Cos[50.19]3 )

𝑀Mid = 24.68 kN. m

0.09 × 52
1 100
𝐹Truss = 0.0171 Cos[50.19]Sin[50.19] ∗ ×
12 0.09 × 52 2 (1 + 0.07335 × Cos[50.19]2) 2
1+ Cos[50.19]Sin[50.19]
12 × 0.0171 (1 + 2 × 0.07335 × Cos[50.19]2)

𝐹Truss = 53.66 kN

Figure 3.21: Case 1b General Layout and Applied Load.

Figure 3.22: Case 1b Member and Hinge Layout.

 36

Figure 3.23: Case 1b Vertical Deflection.

Figure 2.23 shows the vertical deflection at the highlighted nodes. We note that the

deflections are equal. This is expected since the infinitely stiff link causes the direct transfer

of the applied load and therefore the deflection at the top of the arch equals that of the beam.

Figure 3.24: Case 1b Bending Moment Diagram.

 37

Figure 3.25: Case 1b Axial Force Diagram.

Using Figure 3.24 and 3.25 we note that the bending moment and axial force values are

24.33 kN.m and 52.42 kN respectively.

Table 3.3: Case 1b Difference between Derived Formulas and SAP Results.

Studied Parameters Formula Results SAP Results % Difference

δMid 0.01503 mm 0.01509 mm 0.4 %
MMid 24.68 kN.m 24.33 kN.m 1.4 %
FTruss 53.66 kN 52.42 kN 2.4 %

3.2.4 Verification of Case 2a

The final case involves a two-member beam over beam with pinned connections on

both ends. The same numeric values defined in Table 3.1 are applied for Equations 2.7, 2.8,

and 2.9. We obtained the following results.

1
δMid =
3 0.09 ∗ 52
(10 + 22Cos[50.19] − 7Cos[3 × 50.19] + 6Cos[4 × 50.19] + Cos[5 × 50.19]) + Cos[50.19]Sin[50.19] 2
48 12 × 0.0171

100 × 53
× = 1.311 × 10−5 m
48 × 200 × 106 × 0.0171
 38

(−1 − Cos[50.19] + Cos[50.19]2 )

𝑀mid =
1
(1 + 3Cos[50.19] − 3Cos[50.19]2 − 3Cos[50.19] 3 + 3Cos[50.19]4 + Cos[50.19]5 ) + × 1792.54 × Cos[50.19]Sin[50.19]2
12

100 × 5
× = −2.65 kN. m
4

2 × 1792.54 × Sin[2 × 50.19] 100

𝐹truss = ×
(66 + 1792.54)Cos[50.19] − (21 + 1792.54)Cos[3 × 50.19] + 3(10 + 6Cos[4 × 50.19] + Cos[5 × 50.19]) 2

= 63.39 kN

Figure 3.26: Case 2a Member and Hinge Layout.

Figure 3.27: Case 2a Vertical Deflection.

 39

Figure 3.28: Case 2a Bending Moment Diagram.

Figure 3.29: Case 2a Axial Force Diagram.

Using Figure 3.28 and 3.29 we note that the bending moment and axial force values are

2.57 kN.m and 63.44 kN respectively.

 40

Table 3.4: Case 2a Difference between Derived Formulas and SAP Results.

Studied Parameters Formula Results SAP Results % Difference

δMid 0.01311 mm 0.01315 mm 1.0 %
MMid 2.65 kN.m 2.57 kN.m 3.1 %
FTruss 63.39 kN 63.44 kN 0.1 %

Now that all cases have been modeled with SAP and numeric results have been

obtained using the direct stiffness equation, we can compare the results to check the validity

of the direct stiffness equations. We see that the largest difference between results presented

in Tables 3.1, 3.2, and 3.3 is 3.1%. This confirms that our work is verified since the

percentage difference is within an acceptable margin of error.

3.3 Arc of Circle Arch

3.3.1 Definition of Arc of Circle Arch

There are several possible arch configurations that can be modeled and studied. These

include an arc of a circle, parabolic arc, ogee, and depressed arches. The remainder of this

project will consider arc of circle arches. The two main parameters that need to be considered

for this case are the height of the arch above the deck and the width of the arch between the

points of intersection with the deck.

Figure 3.30 illustrates the basic arch bridge graphic used throughout the Wolfram

Mathematics program. We also see the three critical dimensions that need to be defined in

order to create an arc of circle arch. These dimensions include:

 L – This is the entire span of the bridge deck. There are pinned or fixed supports at

either end of the bridge deck.

 41

 LMid – This dimension defines the distance between the intersection of the arch and

bridge deck. In the case of LMid = L, the bridge is a through arch bridge and not a half-

though arch bridge anymore.

 HTop – This is the distance from the middle of the bridge deck to the top of the arch. The

largest value that can be obtained is 0.5× LMid.

Figure 3.30: Half-Through Arch Bridge Dimensions.

3.3.2 Limitations of Arc of Circle Arch

Since an arc of a circle will be used to define the arch for the bridge, there are certain

limitations that arise and need to be considered. These limitations prevent any combinations

of HTop and LMid from working. Equations 3.1 and 3.2 shows the conditions that need to be

satisfied in order to have a valid set of bridge dimensions.

LMid ≥ 2√𝐻𝑇𝑜𝑝 (1 − 𝐻𝑇𝑜𝑝 ) (3.1)

L ≥ LMid (3.2)
 42

CHAPTER 4
PARAMETRIC STUDIES OF ARC OF CIRCLES

4.1 Range of Values Considered

This chapter will cover a comprehensive parametric study using Wolfram Mathematica

to analyze a half-through arch bridge with an arc of circle arch. The first step in this

parametric study is to normalize the parameters that will be varied.

L̂Mid = LMid / 𝐿  This variable ranges from zero to one. When it is equal to zero this means

the arch does not exist. At a value of one, the arch and bridge deck start

and end at the same point. In this case the model becomes a through arch

bridge.

̂𝑇𝑜𝑝 = 𝐻𝑇𝑜𝑝 / LMid  This variable ranges from zero to half since the largest possible arc of
𝐻

a circle is a semicircle with a radius equal to 0.5×LMid.

Applying the previous limitation specified in Equations 3.1 and 3.2 for the normalized

parameters, the values to be considered can be derived. These values will be seen in the

remainder of this section in the summarized tables. The parametric study will include the

following results: arch bending moment, deck bending moment, and deflections at the middle

of the deck.

These results will each be obtained twice, once assuming a hinged connection between

the arch and deck and another assuming a stiff connection. The connection between the arch

and deck is modeled as a rotary spring. At a spring stiffness equal to zero, the connection is

hinged. As the spring stiffness increases the connection also tends to stiffen preventing any

rotation.

Only pinned support at the ends of the arch and deck will be considered for this

parametric study. The flexural and axial stiffness of the arch will be varied for three possible
 43

cases to study the change in load distribution as the arch stiffness increases. EI Arch will be

taken as 0.1, 1, and 10. Also, EA / EI = 1000 for both the arch and deck. The corresponding

EA Arch values will be 100, 1000, and 10000, respectively.

Table 4.1: Numeric Values Assigned for Mathematica Model.

Parameter Numeric Value

L 1
EA Deck 1000
EI Deck 1
EA Ties ∞
P 1

The values in Table 4.1 are used to initialize the Mathematica program in order to

create tables for a parametric study. A total of six study cases are considered as previously

discussed, the obtained results are summarized in the following sections.

These tables can be used to easily identify the optimum values of L̂Mid and 𝐻
̂𝑇𝑜𝑝 in

order to minimize the member bending moments and vertical deflections. Tables 4.2 and 4.3

show the basic layout of the arch bridge for various valid combinations of L̂Mid and 𝐻
̂𝑇𝑜𝑝 .

This is particularly useful in visualizing the shape of the arch, and correlating the physical

shape to its theoretical strength.

 44

Table 4.2: Pinned Arch and Deck – Various Arch Layouts.

Table 4.3: Pinned Arch and Deck – Various Arch Layouts with Deformation.
 45

4.2 Results and Discussions

4.2.1 Study 1: Pinned Arch and Deck - Hinged Connection - EI Arch = 0.1

Figure 4.1: Pinned Arch and Deck - Hinged - EI Arch = 0.1 - Arch Moment.

Table 4.4: Pinned Arch and Deck - Hinged - EI Arch = 0.1 - Arch Moment.
 46

Figure 4.2: Pinned Arch and Deck - Hinged - EI Arch = 0.1 - Deck Moment.

Table 4.5: Pinned Arch and Deck - Hinged - EI Arch = 0.1 - Deck Moment.
 47

Figure 4.3: Pinned Arch and Deck - Hinged - EI Arch = 0.1 - Deflection.

Table 4.6: Pinned Arch and Deck - Hinged - EI Arch = 0.1 - Deflection.
 48

4.2.2 Study 2: Pinned Arch and Deck - Hinged Connection - EI Arch = 1

Figure 4.4: Pinned Arch and Deck - Hinged - EI Arch = 1 - Arch Moment.

Table 4.7: Pinned Arch and Deck - Hinged - EI Arch = 1 - Arch Moment.
 49

Figure 4.5: Pinned Arch and Deck - Hinged - EI Arch = 1 - Deck Moment.

Table 4.8: Pinned Arch and Deck - Hinged - EI Arch = 1 - Deck Moment.
 50

Figure 4.6: Pinned Arch and Deck - Hinged - EI Arch = 1 - Deflection.

Table 4.9: Pinned Arch and Deck - Hinged - EI Arch = 1 - Deflection.

 51

4.2.3 Study 3: Pinned Arch and Deck - Hinged Connection - EI Arch = 10

Figure 4.7: Pinned Arch and Deck - Hinged - EI Arch = 10 - Arch Moment.

Table 4.10: Pinned Arch and Deck - Hinged - EI Arch = 10 - Arch Moment.
 52

Figure 4.8: Pinned Arch and Deck - Hinged - EI Arch = 10 - Deck Moment.

Table 4.11: Pinned Arch and Deck - Hinged - EI Arch = 10 - Deck Moment.
 53

Figure 4.9: Pinned Arch and Deck - Hinged - EI Arch = 10 - Deflection.

Table 4.12: Pinned Arch and Deck - Hinged - EI Arch = 10 - Deflection.

 54

4.2.4 Study 4: Pinned Arch and Deck - Stiff Connection - EI Arch = 0.1

Figure 4.10: Pinned Arch and Deck - Stiff - EI Arch = 0.1 - Arch Moment.

Table 4.13: Pinned Arch and Deck - Stiff - EI Arch = 0.1 - Arch Moment.
 55

Figure 4.11: Pinned Arch and Deck - Stiff - EI Arch = 0.1 - Deck Moment.

Table 4.14: Pinned Arch and Deck - Stiff - EI Arch = 0.1 - Deck Moment.
 56

Figure 4.12: Pinned Arch and Deck - Stiff - EI Arch = 0.1 - Deflection.

Table 4.15: Pinned Arch and Deck - Stiff - EI Arch = 0.1 - Deflection.
 57

4.2.5 Study 5: Pinned Arch and Deck - Stiff Connection - EI Arch = 1

Figure 4.13: Pinned Arch and Deck - Stiff - EI Arch = 1 - Arch Moment.

Table 4.16: Pinned Arch and Deck - Stiff - EI Arch = 1 - Arch Moment.
 58

Figure 4.14: Pinned Arch and Deck - Stiff - EI Arch = 1 - Deck Moment.

Table 4.17: Pinned Arch and Deck - Stiff - EI Arch = 1 - Deck Moment.
 59

Figure 4.15: Pinned Arch and Deck - Stiff - EI Arch = 1 - Deflection.

Table 4.18: Pinned Arch and Deck - Stiff - EI Arch = 1 - Deflection.

 60

4.2.6 Study 6: Pinned Arch and Deck - Stiff Connection - EI Arch = 1

Figure 4.16: Pinned Arch and Deck - Stiff - EI Arch = 10 - Arch Moment.

Table 4.19: Pinned Arch and Deck - Stiff - EI Arch = 10 - Arch Moment.
 61

Figure 4.17: Pinned Arch and Deck - Stiff - EI Arch = 10 - Deck Moment.

Table 4.20: Pinned Arch and Deck - Stiff - EI Arch = 10 - Deck Moment.
 62

Figure 4.18: Pinned Arch and Deck - Stiff - EI Arch = 10 - Deflection.

Table 4.21: Pinned Arch and Deck - Stiff - EI Arch = 10 - Deflection.

 63

4.3 Discussions of Parametric Study

Following the parametric study, a series of discussions is required to interpret the

results and deduce what arch dimensions and properties are best for an overall optimized

structure. The first step is to identify the values of L̂Mid and 𝐻

̂𝑇𝑜𝑝 that correspond to the

smallest moment and deflection coefficients. Upon observation these values are found to be

0.6 and 0.1, respectively. Table 4.22 summarizes the corresponding coefficients obtained

from Tables 4.4 to 4.21.

Table 4.22: Various Parametric Study Coefficients from Study 1.

Table Component EI Arch Coefficients

Table 4.4 Arch Moment - Hinged 0.1 0.006149
Table 4.5 Deck Moment - Hinged 0.1 0.050605
Table 4.6 Deflection - Hinged 0.1 0.038509
Table 4.7 Arch Moment - Hinged 1 0.018063
Table 4.8 Deck Moment - Hinged 1 0.015591
Table 4.9 Deflection - Hinged 1 0.007462
Table 4.10 Arch Moment - Hinged 10 0.026212
Table 4.11 Deck Moment - Hinged 10 0.002320
Table 4.12 Deflection - Hinged 10 0.000919
Table 4.13 Arch Moment - Stiff 0.1 0.005799
Table 4.14 Deck Moment - Stiff 0.1 0.050873
Table 4.15 Deflection - Stiff 0.1 0.038455
Table 4.16 Arch Moment - Stiff 1 0.017645
Table 4.17 Deck Moment - Stiff 1 0.015962
Table 4.18 Deflection - Stiff 1 0.007448
Table 4.19 Arch Moment - Stiff 10 0.026128
Table 4.20 Deck Moment - Stiff 10 0.002400
Table 4.21 Deflection - Stiff 10 0.000919
 64

We immediately notice that as the stiffness of the arch increases the normalized

moment coefficient also increases. This true for both cases of arch to deck connection. It is

also worth noting that when the connection becomes stiff the arch moment coefficient is

smaller than the previous case. This is due to some moment being transferred into the deck.

An inverse observation is made for the deck normalized moment and deflection

coefficients. These values decrease as the arch stiffness is increased. This is an expected

behavior since as the arch resists more load, the deck load is reduced.

Minimal change in deflection is noted when the arch to deck connection is changed

from hinged to stiff. While the deck moment coefficients are seen to increase slightly when

the connection is stiffened.

Based on these observations it can be concluded that the best possible configuration is

dependent on the design of the arch and deck. There are various ways of achieving different

load distributions, and based on the required design an adequate analysis can be provided.
 65

CHAPTER 5
CONCLUSIONS AND RECOMMENDATIONS

5.1 Conclusions

This project has broadened our knowledge towards the behavior of arches. Starting with

simple examples and progressing towards more complex cases allowed us to build an

understanding and appreciation for the Direct Stiffness Method and its systematic approach to

structures.

Solving the initial three cases using manual calculations and SAP was an important step

in verifying the credibility of our work and significant conclusions were made. The first

being that a negligible difference in optimal angle is noticed between double pinned and

pinned - roller supports. This is useful in simplifying analysis calculations even though in

reality a roller support is used. We also noticed that to minimize deflection, an angle between

45 to 60 degrees should be used regardless if the arch is a truss or frame system. Parametric

studies involving about 3600 cases were done for the arc of a circle arch case. The results

provide better understanding as a basis to design half-through arch bridges.

5.2 Recommendations

The potential improvements for this project are numerous. Additional research and

analysis could lead to further understanding the parameters of an arch bridge and how to

optimize them to a maximum. Potentials recommendations include:

 Expand arch configuration to other cases such as arcs of parabolas and ogee arches.

 Studying various distributed vertical loads and other real-world load simulations.

 Use derived analysis results to design member sections based on different materials.
 66

LIST OF REFERENCES

[1] Belevicius, R., Juozapaitis, A., & Rusakevičius, D. (2018). Parameter study on weight

minimization of network arch bridges. Periodica Polytechnica, Civil Engineering,

62(1), 48-55. doi: 10.3311/PPci.10036

[2] Chróğcielewski, J., Malinowski, M., & Miğkiewicz, M. (2008). Arch bridges under test

loadings. Journal of KONES Powertrain and Transport, 1(1), 29-38.

doi:10.1.1.836.7379

[3] Fares, N. (2012). Practical approximate analysis of beams and frames [PDF document].

Reston, VA: American Society of Civil Engineers. Retrieved from Lecture Notes on

Mechanics 1 Online: https://ascelibrary.org/doi/pdf/10.1061/9780784412220.fm

[4] Henderson, J. (Photographer). (2008). Bayonne bridge [Photograph]. New York City,

New Jersey.

[5] J.W.C., A. (Photographer). (2008). Sydney harbour bridge [Photograph]. Sydney,

Australia.

[6] Magalhães, F., Cunha, Á., & Caetano, E. (2008). Dynamic monitoring of a long span arch

bridge. Engineering Structures, 30(11), 3034-3044. doi:

10.1016/j.engstruct.2008.04.020

[7] Primary Pros and Cons of Arch Bridges (2017, September 7). In Green Garage Blog.

Retrieved May 10, 2018, from https://greengarageblog.org/14-primary-pros-and-cons-

of-arch-bridges
 67

[8] Salamak, M., Radziecki, A., Lazinski, P., & Pradelok, S. (2013, October). Analysis of

The Results from The Load Testing of Steel Through Arch Bridges. 7th International

Conference on Arch Bridges, 187-194. doi:10.13140/2.1.4962.8161

[9] SAP2000 (Version 19.2.0) [Computer Software]. Berkley, California: Computers and

Structures, Incorporation (CSI).

[10] Wolfram Mathematica 11 (Version 11.0.1) [Computer Software]. Champaign, IL:

Wolfram Research.