Beruflich Dokumente
Kultur Dokumente
Bridges
By
Ezz El-Din Kamel Mohamed Ali
Engineer:
Date of Birth:
Nationality:
E-mail:
Phone:
Address:
Registration Date:
Awarding Date:
Degree:
Department:
Supervisors:
Examiners:
Title of Thesis:
AERODYNAMIC STABILITY OF ULTRA-LONG SPAN CABLE SUSPENSION
BRIDGE
Key Words:
Deck shape determines the aerodynamic forces acting on bridges. The present study
focuses on minimizing aerodynamic forces by optimizing deck shape. Computational
fluid dynamic (CFD) is an effective numerical tool for simulating flow around bridge
deck, which was verified in literature. In this study six deck cross sections were studied
and their generated aerodynamic forces were compared with forces acting on Sutong
Bridge Also, stability and structural analysis for ultra-span bridge performed using cablesuspension bridge system
ACKNOWLEDGMENTS
I would like to express my most sincere gratitude and appreciation to my advisors, Prof. Ahmed
Mahmoud Saleh, Prof. Walid Abdel Latif and Prof. Eehab Khalil for thier guidance,
encouragement, and patience throughout the completion my thesis work.
I owe my deepest gratitude to Assoc. Prof Basman Elhadidy, support to develop Ansys models
and revise the aerodynamic part of the study.
I know that I could not have completed this project without the support of my family. Not only
did you both offer me loads of moral support, but you helped me finding the strength to continue
my work here and helped me come to the decision, once and for all.
And of course, I want to thank all of those whose love makes the difference in my life. To my
family, for being the light, joy, and fuel of my existence and for being my constant inspiration
and support.
Abstract
Suspension bridges are longest span structures, however, the major problem of those structures
are its stability under wind loads. The stability of those bridges depends majorly on aerodynamic
forces and structural supporting system; this study worked on those two major axes.
Aerodynamic axis was to minimize aerodynamic forces by optimizing deck shape.
Computational fluid dynamic (CFD) is an effective numerical tool for simulating flow around
bridge deck, which was verified in the literature. six deck cross sections were studied three
elliptical deck shapes and three closed trusses-; their generated aerodynamic forces were
compared with forces acting on Sutong Bridge and the best one was modelled as unsteady- state
model for calculating flutter derivatives. The elliptical deck shape is found to generate the least
static pressure and aerodynamic forces relative to the box and closed truss shapes; that proves, as
deck shape get smoother aerodynamic forces generated by the deck reduces
Structural axis was to improve the stability of cable supporting system. A combination of threedimensional suspension cable and cable-stay systems studied. Catenary theory was improved for
determining three-dimensional suspension cable profile. Three models suspension bridge and
eight models cable-stayed suspension cable were studied. The use of three-dimensional
suspension cable improved critical wind speed.
ii
Table of Contents
ACKNOWLEDGMENTS ............................................................................................................... i
Abstract ........................................................................................................................................... ii
Table of Contents ........................................................................................................................... iii
List of Figures ............................................................................................................................... vii
List of Tables ............................................................................................................................... xiv
Chapter 1 ......................................................................................................................................... 1
Introduction ..................................................................................................................................... 1
1.1.
Introduction ...................................................................................................................... 1
vi
List of Figures
Figure (1.1) Rope bridge. ................................................................................................................ 2
Figure (1.2) Inca bridge .................................................................................................................. 2
Figure (1.3) Jacob's Creek Bridge.................................................................................................. 3
Figure (1.4) Winch Bridge (1830) .................................................................................................. 4
Figure (1.5) The Szchenyi Chain Bridge ...................................................................................... 5
Figure (1.6) Brooklyn Bridge (1883). ............................................................................................. 5
Figure (1.7) Brooklyn Bridge Tower .............................................................................................. 6
Figure (1.8) Tacoma Bridge cross section. ..................................................................................... 7
Figure (1.9) Tacoma Bridge view. .................................................................................................. 7
Figure (1.10) Sutong Bridge ........................................................................................................... 8
Figure (1.11) Span arrangements. ................................................................................................... 8
Figure (1.12) Sutong Bridge Cross section. .................................................................................... 9
Figure (1.13) Sutong Bridge Pylon. ................................................................................................ 9
Figure (1.14) The Rion-Antirion Bridge. ...................................................................................... 10
Figure (1.15) The Rion-Antirion Bridge spans arrangements. ..................................................... 10
Figure (1.16) The Rion-Antirion Bridge Cross section. ............................................................... 10
Figure (1.17) The Rion-Antirion Bridge Cross section details. .................................................... 11
Figure (1.20) Proposed Messina bridge cross sections ................................................................. 13
Figure (1. 21) Proposed Gibraltar Bridge ..................................................................................... 14
Figure (2.1) Da Vinci helicopter ................................................................................................... 16
Figure (2.2) Inclined Plate Forces Sketch. .................................................................................... 17
Figure (2.3) Boundary layer (thesis ref. 7). .................................................................................. 18
Figure (2.4) Reynolds number effect on drag forces (thesis ref. 7). ............................................. 19
Figure (2.5) Reynolds number effect on drag forces. ................................................................... 19
Figure (2.6) Flow separation sketch for bluff body ...................................................................... 20
Figure (2.7) B/D effect for the bluff body with sharp edges. ....................................................... 20
Figure (2.8) Tacoma Bridge Oscillating. ...................................................................................... 21
Figure (2.9) Tacoma Oscillation (torsional mode shape) ............................................................. 22
Figure (2.10) Tacoma collapse. .................................................................................................... 22
Figure (2.11) Tacoma after collapsing. ......................................................................................... 23
Figure (2.12) Collar's aeroelastic triangle. .................................................................................... 24
Figure (2.13) Aerodynamic forces acting on the bridge with angle of attack ........................... 26
Figure (2.14) Deck supporting sketch for aeroelastic wind tunnel testing (Bridge aeroelasticity).
....................................................................................................................................................... 27
Figure (2. 15) Full-scale model ..................................................................................................... 28
Figure (2. 16) Degrees of freedom for bridge deck or airfoil. ...................................................... 29
Figure (2.17) Movement considered for the thin plate. ................................................................ 30
Figure (2.18) Real and imaginary parts of the Theodorsen function ............................................ 31
vii
Figure (2.19) Comparison between measured critical flutter speed and Selberg prediction ........ 33
Figure (2.20) CFD simulation methodology................................................................................. 35
Figure (2.21) A rectangular box solution domain (LxD) .............................................................. 36
Figure (2. 22) cross section of the bridge deck and the sign convention for force ....................... 38
Figure (2. 23) Tested Setup in Wind Tunnel with 0o wind ........................................................... 38
Figure (2. 24) variation of mean drag coefficient with wind angle. ............................................. 39
Figure (2. 25) variation of mean lift coefficient with wind angle. ................................................ 40
Figure (2. 26) variation of mean Moment coefficient with wind angle. ....................................... 41
Figure (2. 27) Domain Decomposition ......................................................................................... 42
Figure (2. 28) Geometry of studied sections used in the present study ........................................ 43
Figure (3. 1) Drag coefficient for a truss section and streamlined box section. ........................... 45
Figure (3. 2) Influence of streaming box section on drag forces .................................................. 46
Figure (3. 3) Drag forces for symmetric and non-symmetric shapes ........................................... 46
Figure (3. 4) Lift forces for the symmetric and non-symmetric airfoil. ....................................... 47
Figure (3. 5) Width over depth ratio effect. .................................................................................. 47
Figure (3. 6) streamlining effect on drag forces. .......................................................................... 47
Figure (3. 7) Proposed DECK 1 Cross section. ............................................................................ 48
Figure (3. 8) Proposed DECK 2 Cross section. ............................................................................ 48
Figure (3. 9) Proposed DECK 3 Cross section ............................................................................. 49
Figure (3. 10) Proposed DECK 4 Cross section ........................................................................... 49
Figure (3. 11) Proposed DECK 5 Cross section ........................................................................... 49
Figure (3. 12) Proposed DECK 6 Cross section ........................................................................... 50
Figure (3. 13) Sutong bridge deck ................................................................................................ 50
Figure (3. 15) CFD model for Proposed Deck 1. .......................................................................... 51
Figure (3. 16) CFD model for Proposed Deck 2. .......................................................................... 51
Figure (3. 17) CFD model for Proposed Deck 3. .......................................................................... 52
Figure (3. 18) CFD model for Proposed Deck 4. .......................................................................... 52
Figure (3. 19) CFD model for Proposed Deck 5. .......................................................................... 53
Figure (3. 20) CFD model for Proposed Deck 6. .......................................................................... 53
Figure (3. 21) Shows drag coefficient for studied deck shaped. ................................................... 54
Figure (3. 22) Shows Domain Decomposition ............................................................................. 56
Figure (3. 23) CFD model for determining flutter derivatives. .................................................... 57
Figure (4. 1) Deck shape 1 streamline (3D view) ......................................................................... 59
Figure (4. 2) Deck shape 1air streamlines (2D view). .................................................................. 60
Figure (4. 3) Deck shape 1 air flow vectors (2D view) ................................................................ 60
Figure (4. 4) Deck shape 1 air flow vectors (3D view) ................................................................ 61
Figure (4. 5) Deck shape 1 Pressure distribution around. ............................................................. 61
Figure (4. 6) Deck shape 1 velocity distribution around. ............................................................. 62
viii
Figure (4. 45) Air velocity distribution around deck under forced vertical motion at 0.15 second
....................................................................................................................................................... 80
Figure (4. 46) Air velocity distribution around deck under forced vertical motion at 0.20 second
....................................................................................................................................................... 80
Figure (4. 47) Air velocity distribution around deck under forced vertical motion at 0.25 second
....................................................................................................................................................... 81
Figure (4. 48) Air velocity distribution around deck under forced vertical motion at 0.30 second
....................................................................................................................................................... 81
Figure (4. 49) Air velocity distribution around deck under forced vertical motion at 0.35 second
....................................................................................................................................................... 81
Figure (4. 50) Air velocity distribution around deck under forced vertical motion at 0.40 second
....................................................................................................................................................... 82
Figure (4. 51) Air velocity distribution around deck under forced vertical motion at 0.45 second
....................................................................................................................................................... 82
Figure (4. 52) Air velocity distribution around deck under forced vertical motion at 0.50 second
....................................................................................................................................................... 82
Figure (4. 53) Forced rotational motion at time 0.5 second.......................................................... 83
Figure (4. 54) Forced rotational motion at time 1.0 second.......................................................... 83
Figure (4. 55) Forced rotational motion at time 1.5 second.......................................................... 83
Figure (4. 56) Forced rotational motion at time 2.0 second.......................................................... 84
Figure (4. 57) Forced rotational motion at time 2.5 second.......................................................... 84
Figure (4. 58) Pressure distribution around deck under forced Torsional motion at 0.05 second 84
Figure (4. 59) Pressure distribution around deck under forced Torsional motion at 0.10 second 85
Figure (4. 60) Pressure distribution around deck under forced Torsional motion at 0.15 second 85
Figure (4. 61) Pressure distribution around deck under forced Torsional motion at 0.20 second 86
Figure (4. 62) Pressure distribution around deck under forced Torsional motion at 0.25 second 86
Figure (4. 63) Pressure distribution around deck under forced Torsional motion at 0.30 second 87
Figure (4. 64) Pressure distribution around deck under forced Torsional motion at 0.35 second 87
Figure (4. 65) Pressure distribution around deck under forced Torsional motion at 0.40 second 88
Figure (4. 66) Pressure distribution around deck under forced Torsional motion at 0.45 second 88
Figure (4. 67) Pressure distribution around deck under forced Torsional motion at 0.50 ............ 89
Figure (4. 68) Air velocity distribution around deck under forced Torsional motion at 0.05
second ........................................................................................................................................... 89
Figure (4. 69) Air velocity distribution around deck under forced Torsional motion at 0.10
second ........................................................................................................................................... 90
Figure (4. 70) Air velocity distribution around deck under forced Torsional motion at 0.15
second ........................................................................................................................................... 90
Figure (4. 71) Air velocity distribution around deck under forced Torsional motion at 0.20
second ........................................................................................................................................... 91
Figure (4. 72) Air velocity distribution around deck under forced Torsional motion at 0.25
second ........................................................................................................................................... 91
Figure (4. 73) Air velocity distribution around deck under forced Torsional motion at 0.30
second ........................................................................................................................................... 92
Figure (4. 74) Air velocity distribution around deck under forced Torsional motion at 0.35
second ........................................................................................................................................... 92
Figure (4. 75) Air velocity distribution around deck under forced Torsional motion at 0.40
second ........................................................................................................................................... 93
Figure (4. 76) Air velocity distribution around deck under forced Torsional motion at 0.45
second ........................................................................................................................................... 93
Figure (4. 77) Air velocity distribution around deck under forced Torsional motion at 0.50
second ........................................................................................................................................... 94
Figure (4. 78) forced non-dimensional vertical bending (h/B =0.05) simulated and Corresponding
CL time history (U/fB =6). ............................................................................................................ 94
Figure (4. 79) forced non-dimensional vertical bending (h/B =0.05) simulated and Corresponding
CM time history (U/fB =6). ........................................................................................................... 95
Figure (4. 80) forced non-dimensional torsional bending 0 simulated and corresponding CM time
history (U/fB =6). ......................................................................................................................... 95
Figure (4. 81) Flutter Derivatives H1, H2, H3, and H4 ................................................................... 96
Figure (4. 82) Flutter Derivatives A1, A2, A3, and A4 ................................................................... 96
Figure (5. 1) Wing statical system ................................................................................................ 99
Figure (5. 2) Proposed Messina Bridge ........................................................................................ 99
Figure (5. 3) Akashi Kayoka bridge. .......................................................................................... 100
Figure (5. 4) Ellipse Characteristics............................................................................................ 100
Figure (5. 5) Proposed Elliptical Shape ...................................................................................... 101
Figure (5. 6) typical vertical bracing for one span ...................................................................... 101
Figure (5. 7) Separated Boxes..................................................................................................... 101
Figure (5. 8) Design truss section process .................................................................................. 102
Figure (5. 9) Proposed Box section............................................................................................. 102
Figure (5. 10) Load Model No. 1 ................................................................................................ 104
Figure (5. 11) Load Model No.2 (single axis) ............................................................................ 104
Figure (5. 12) Load Model 71 and characteristic values for vertical loads. ............................... 105
Figure (5. 13) four spans model using SAP2000. ....................................................................... 107
Figure (5. 14) Deck model using SAP 2000. .............................................................................. 107
Figure (5. 15) Resale assignment on top bracing. ....................................................................... 107
Figure (5. 16) box section model in SAP2000............................................................................ 108
Figure (5. 17) Simple Beam load and displacement ................................................................... 109
Figure (5. 18) Twisting of a beam .............................................................................................. 109
Figure (5. 19) Deck model sketch for calculating torsional stiffness. ........................................ 110
xi
Figure (6. 15) deformed shape under dead loads a. Model 3 and b. Model 4. ........................... 148
Figure (6. 16) Hybrid Bridge bending moment M2-2 ................................................................ 149
Figure (6. 17) Hybrid Bridge bending moment M3-3 ................................................................ 150
Figure (6. 18) Hybrid Bridge Axial force for the Pylon ............................................................. 151
Figure (6. 19) Hybrid Bridge Axial force for the main cable ..................................................... 152
Figure (6. 20) Hybrid Bridge bending-deformed shape.............................................................. 153
Figure (6. 21) Hybrid Bridge bending mode shape .................................................................... 154
Figure (6. 22) Hybrid Bridge Torsion mode shape ..................................................................... 155
Figure (6. 23) Hybrid Bridge Torsion mode shape ..................................................................... 156
xiii
List of Tables
Table (1. 1) Messina bridge data................................................................................................... 12
Table (2. 1) Comparison of simulation and experimental ............................................................ 34
Table (2. 2) recommended mesh height for different shapes........................................................ 42
Table (2. 3) studied models mesh characteristics. ....................................................................... 43
Table (3. 1) Studied deck shapes .................................................................................................. 50
Table (3. 2) recommended body fitted grid height ....................................................................... 53
Table (3. 3) Domain decomposition factors ................................................................................. 56
Table (3. 4) Reduced velocity and forced motion amplitudes. ..................................................... 58
Table (5. 1) Reduced velocity and forced motion amplitudes. ..................................................... 75
Table (5. 2) Proposed deck properties .......................................................................................... 96
Table (5. 1) static aerodynamic forces acting on the bridge Deck .............................................. 106
Table (5. 2) Proposed deck properties ........................................................................................ 110
Table (5. 3) studied cables configurations .................................................................................. 123
Table (5. 4) studied parameters. .................................................................................................. 131
Table (6. 1) natural frequencies and critical wind speed ............................................................ 132
Table (6. 2) Models Major results ............................................................................................. 132
Table (6. 4) Deck deformations .................................................................................................. 144
Table (6. 5) Main cable deformations ......................................................................................... 144
Table (6. 6)) natural frequency (Hz) and critical wind speed (m/sec). ....................................... 157
xiv
Chapter 1
Introduction
1.1. Introduction
Cable-supported bridges are most important bridge system used for crossing large spans and
connecting great distances. The most common types are cable-stayed bridges and suspension
bridges. Even though their origins and evolution processes are different, their analysis methods
are very similar to the present. Thus, the term, cable supported bridge is often used to refer to
both types of structures.
The concept of suspension bridges was used in bridges more than two hundred years ago, and it
is almost the same nowadays. However, the span length of the large modern bridges is one
hundred times longer than that of the first known bridge. On the contrary, the history of cablestayed bridges is quite recent than suspension bridges. It was first introduced at the beginning of
the 20th century.
The following sections of this chapter will discuss long bridges over time and many proposed
ambitious projects underway.
pedestrians. As a result, the Incan Empire did not use trundled modes of transport. Bridges were
a major part of Incas road system and were a significant engineering innovation.
In the 18th century, it was the first introduction of the suspension bridges in simple form as it
does appear nowadays. The following most famous cable bridges will be discussed.
Jacob's Creek Bridge (1801, demolished 1833)
James Finney designed and built the first steel suspension bridge with 21 m central span
and 3.51 width over the Jacobs Creek in 1801, figure (1.3). The bridge was the first iron-chained
suspension bridge built in the United States, located south of Mount Pleasant, Pennsylvania.
Tacoma Bridge collapse changed the world view of cable bridges analysis and design.
Aeroelasticity was firstly introduced after deep analysis for Tacoma collapse as would be
explained later.
Sutong Bridge in China
Sutong Bridge was the longest cable-stay bridge with 1088 meters central span and a 300-meter
pylon height. The bridge crosses Nantong Yangtze River in China and completed in 2007 figure
(1.10). The total length of the cable supported part was 2088 meters, and the pylon height was
300 meter.
concrete of C50 grade. The upper part is a steel part fastened to concrete using shear studs. The
steel part used for anchoring stays cables, figure (1.13)
The RioAntirio Bridge is the longest multi-span cable stay bridges; the bridge crosses
the Gulf of Corinth near Patras, Greece. The bridge consists of five spans with four pylons.
Spans are 286-560-560-560-286 meters with a total length 2800 meters, figure (1.14 and 1.15).
The girder carries four dual-traffic lanes and an emergency lane for a pedestrian walkway
with 27.20 meteres total width, figure (1.16). The deck is composite steel-concrete structure; 2535 cm concrete slab connected to twin longitudinal steel I-girders, 2.20 meters high, figure
(1.17).
10
11
3300 m
382.6
60.4 m
4 x 1.24m
75 m/sec
7.1 magnitude
200 years
6.1 bn
Different cross-sections were studied, figure (1.20) and section (C) was chosen to be the most
suitable for the bridge.
Gibraltar Bridge
The longest proposed bridge in the world. The bridge will connect Morocco, Africa and Spain,
Europe. Engineers proposed several ingenious design for a bridge or tunnel across the Strait of
Gibraltar; that will join the continents of Europe and Africa.
In 1996 Prof. T Y Lin, designed a hybrid stayed suspension bridge after numerous Gibraltar
Bridge alternatives had been exhausted, including classic suspension bridges and cable-stiffened
suspension bridges, figure (1.21). The hybrid stayed suspension bridge was the most suitable
alternative due to its better rigidity and better aerodynamics.
(a)
12
(b)
(c)
Figure (1.18) Proposed Messina bridge cross sections
13
14
15
Chapter 2
Literature Review
16
Equation (2. 1)
2.1. Aerodynamic.
Until the 7th November 1940, Cable bridges were mainly designed for traditional static
and live loads considering the static effect of wind. At this date, Tacoma Bridge collapse;
Tacoma Bridge was the longest suspension bridge with a central span 853 m.
Tacomas collapse caused an immediate commotion and widespread unrest. In the words of Scott
how can an apparently well-designed bridge using the latest technology to withstand winds of
161 km/h and the static horizontal pressure of winds 146 kg/m2 subjected to winds at half the
speed and pressures around one-sixth of the design value?
The major problems of cable-supported bridges under wind are those associated with excess
deflection, which caused by oscillatory instability or by random response due to the action of
wind gusts.
In the case of cable bridges, the bridge is very flexible and the system effect majorly by the wind.
It is primarily the recurrence of large deflections and truly aeroelastic behavior of the bridge that
has set the study of cable bridges apart as a particular subset of the wind problems of cablesupported structures.
Cable-supported bridges are classified as bluff bodies. The flow separation and re-attachment
characterize this type of body as a source of excitation that may lead to aeroelastic instabilities.
The following section discusses aerodynamics concepts as a boundary layer for static and
oscillating bodies.
17
18
Figure (2.4) Reynolds number effect on drag forces (thesis ref. 7).
19
characteristic of the boundary layer. The second is a geometric factor characterizing the bluffness
of the body (B/D), where B and Dare the width and the depth of the body, respectively
Figure (2.7) illustrate the geometry factor, where flow separation is sketched for two different
bluff bodies B/D = 1, B/D= 2 B/D = 7. The modification of bluff body sharpness into rounded
corners or streamlined aerodynamic appendages is an economical method to reduce the flow
separation and the resulting aeroelastic instabilities.
Figure (2.7) B/D effect for the bluff body with sharp edges.
20
21
22
Tacoma Bridge was well designed and constructed. However, it could resist all static
forces neglecting the dynamic effect of wind. The wind caused undulation that led the
bridge failure.
The bridge was flexible, and the bridge was not able to absorb or damp the dynamic
forces that caused oscillation. This oscillation led to the bridge failure.
The vertical oscillation caused no structural damage. However, the torsional oscillation
stressed the cables that led to the failure of suspenders though central span, then the
progressive collapse of the bridge occurred.
They recommended more experimental studies for suspension bridges to determine the
aerodynamic forces acting on it and bridge behavior.
Von Karman, a known aeronautical engineer, studied the Tacoma bridge failure and proposed
that the motion that caused the failure was due to vortices. He explained that when an air flow
passes a bluff body, periodic shedding of air vortices occurs. Those vortices caused central span
oscillation and violent twist until the bridge collapse.
K Yusuf Billah and Robert H. Scanlan proposed that the reason for the collapse was the
aerodynamic flutter; that the change of the lift and drag forces along the span of the bridge
during oscillation. The torsional degree of freedom allowed torsional oscillation. This oscillation
occurred due to wind load. As the deck rotated the wind forces acting on the bridge deck
23
changed and the wind forces pushed the bridge deck in the opposite direction. As the bridge
rotated, forces acting on the bridge increased, the bridge failed; the bridge was free to rotate, and
this rotation led to the bridge collapse.
The studies and explanations for Tacoma Bridge collapse ensured that the cable bridges are
flexible and are not able to absorb aerodynamic forces. The deck oscillated due to vortices
formation; as the bridge oscillated, the forces acting on the bridge changed. This relation
between bridge stiffness and aerodynamic forces for a flexible structure as cable bridges is a
complicated. Aeroelasticity is the science that discusses such relationships.
2.3.Aeroelasticity
The air flow around a bridge causes deformation and changes the wind load acting on the
structure. As the wind pressure increases, bridge deformation increases; the bridge acts as an
elastic structure. The phenomena discussing this relation is called aeroelastic, and the science is
aeroelasticity. Aeroelasticity is the science that discusses the relation between elastic structures,
and dynamic forces results from fluid movement around the structure, figure (2.12) colars
aeroelastic triangle
bridge collapse. However, they might make the bridge uncomfortable for walking or limit traffic
circulation
Buffeting
An aeroelastic vibration phenomenon generated by turbulent wind flow; that occurs when the
wind gusts of different average wind velocity. This phenomenon does not cause instability like
vortex shedding, but both affect the service lifetime of the bridge.
Flutter
The interaction between bridge and wind flow generates an incremental vibration; vibration
amplitude increases with each cycle. Once vibration starts, aeroelastic instabilities starts, and it
occasionally causes bridge collapse. This aeroelastic oscillation vibration can occur at a uniform
flow without external disturbance.
Cable vibration
Cables of cable stay bridges also interact with air fluid. This interaction generates cable
vibrations that may occur in different velocity ranges and is influenced by the associated rain.
The deformations of cables depend on its geometry, stiffness, and mass. This phenomenon does
not cause cable rupture. However, some events of great amplitude and frequency have happened
so that correctors devices had to be installed.
All the aeroelastic phenomena discussed before are critical. However, the flutter is the most
critical one because it is the only one that occasionally may cause a bridge collapse.
2.2. Flutter
Two major approaches that facilitate the study of the behavior of cable-supported bridges under
wind loads. The rst one is the experimental aeroelasticity method, which is based on windtunnel testing with reduced scale models. The other is the analytical approach that is based on
aeroelasticity analysis, applied already in aeronautical engineering. The flutter instability may be
done by considering a single type of vibration, or study of the combination of various vibration
modes that may occur due to the interaction of different component within each mode.
Considering different modes provides safer and more accurate results for bridges whose span
length is between 1500 and 2000 meters. The following section discusses experimental and
numerical methods for studying flutter phenomena for cable bridges.
25
Figure (2.13) Aerodynamic forces acting on the bridge with angle of attack
Equation (2. 3)
Where:
D is drag force;
L is lift force;
M is pitching moment;
is air density;
B is deck length;
U is uniform wind speed.
This test is performed for different wind angle of attack, and the aerodynamic forces are
measured for each angle. These forces will be used in calculating aerodynamic coefficients that
are non-dimensional coefficient representing aerodynamic force the recommendation scale for
this test varies from 1:80 to 1:40 for a bridge under design or study.
This test simulates bridge deck as possible as to the original type. Bridge deck model in this test
has three degrees of freedom, v, w and figure (2.15). For simulating that displacement, a
dimensional analysis should be carried out for determining mass, mass model and stiffnesses of
elastic springs which as supports for the deck model, figure (2.14). Figure (2.15) displays a
sketch of aeroelastic forces. Those forces were formulated as Equations (2.4, 2.5 and 2.6).
Aeroelastic forces measured during the test as time history and flutter derivatives can be easily
extracted from those equations.
Figure (2.14) Deck supporting sketch for aeroelastic wind tunnel testing.
(
)
Equation (2. 4)
)
Equation (2. 5)
27
)
Equation (2. 6)
Where:
28
29
,(
Equation (2. 7)
,(
(
)(
-
)*
Equation (2. 8)
Where U is the average of the acting wind speed, the air density and F and G are the real and
imaginary parts of the Theodorsen function respectively
30
( )
( )
( )
( )
( )
( )
( )
( )
( )
( )
( )
Equation (2. 9)
Where: H and K are the modified Hunkel and Bessel functions. The real part F(k) and the
imaginary G(k) of C(k)
)(
)(
In this form, there are six aerodynamic derivatives. Later on, they improved those equations that
involve eight aerodynamic derivatives as follows:
31
)(
)(
Where:
and
are flutter derivatives
The lift force may be expressed in a classical expression per unit length of span for small wind
angle of attack:
( )
. Those terms
Terms
or
are thus analogous to lift coefficient derivatives
should be referred to as motional derivatives, however, and they go over into steady-state
derivatives, such as
for K equals zero. Experimentally, aerodynamic coefficients can
( )(
( )
Where:
Figure (2.19) Comparison between measured critical flutter speed and Selberg prediction
Computational fluid dynamics (CFD) is a numerical simulation of fluid flow, heat transfer, mass
transfer, chemical reactions, and related phenomena. Mathematical algorism flows governing
equation for predicting the studied phenomena behavior. CFD helps in simulating the interaction
between air and fluid by defining a boundary layer between different medias.
CFD has made impressive progress in the past decade and has evolved into a promising design
tool for the development of Aerodynamic structures. CFD is cheaper than tunnel test; using CFD
models. It is easy to calculate aerodynamic forces acting on any shape with reasonable accuracy.
The parametric study became more competitive, saving the cost of tunnel tests trails. Only
verifying the optimum aerodynamic shape with a prototype model to verify the CFD results is
required. Table (2.1) shows a comparison between CFD simulation and experimental work.
Table (2. 1) Comparison of simulation and experimental
Simulation (CFD)
Cost
Cheap
Time
short
Scale
Any
Information
All
Repeatable
Yes
Safety
Yes
Experimental
Expensive
Long
Small/Middle
Measured Points
Some
Some Dangerous
Momentum Equation:
Where:
34
Energy Equation:
35
36
37
Figure (2. 22) cross section of the bridge deck and the sign convention for force
This section was tested in a wind tunnel; the wind direction angle varied from -15 to 15 degrees,
figure (2.23). The wind tunnel test has 1800 cm length, 250 cm width, and 180 cm height.
38
In wind tunnel test, aerodynamic forces are computed using the measured pressure. The forces
calculated for X and Y directions; those forces were resolved in drag force acting along the
direction of wind and lift forces acting perpendicular to the direction of the wind.
Where and are mean forces coefficients along x and y axis; and were computed using
measured pressures. B is the characteristic length and was assumed to be the height of the bridge
section/ the dimension normal to the flow. Drag forces are the forces acting along the wind
direction, and the lift forces acting perpendicular to wind direction. Fx and FY were measured,
then drag and lift forces computed based on wind direction angle. Aerodynamic coefficients
calculated as following
Figure (2. 24) variation of mean drag coefficient with wind angle.
39
Figure (2. 25) variation of mean lift coefficient with wind angle.
40
Figure (2. 26) variation of mean Moment coefficient with wind angle.
From the comparisons, it was found that RM2006 gives more realistic results than other models;
the CD and CL results match very well with the values obtained based on wind tunnel tests, in the
opposite CM does not match.
41
U=V=0
W2
W4
W5
W6
V1
V3
Bridge Section
V5
V6 Rigid Boundary Layer Mesh
V4
Dynamic Mesh Region
P=0
W3
Outlet Boundary
W1
V2
Static Mesh Region
U=V=0
Computational Model
Re
Notes
0.02D
250-8e-05
D is Diameter
0.028B
2.2e4
B is width
0.02B
2e4
0.021B
1.1e5
B is width
0.019B
1.5e5
B is width
Zhu
0.01B
300
B is width
Cao
<0.04B
1.5e5
B is width
height
Vario
Five bridge deck cross sections were studied, figure (2.28). Table (2.3) summarized
computational and mesh characteristics respectively. Lin. Huang, et al. compared results and
wind tunnel analysis and good agreement were found. They concluded that their proposed
computational method is effective can be considered as a practical tool for simulating
aerodynamic flutter derivatives.
42
Figure (2. 28) Geometry of studied sections used in the present study
Table (2. 3) studied models mesh characteristics.
Model
Total mesh
Quadrangular
mesh
P
38192
7213
GA
26300
4804
GB
35037
13541
SA
45939
13597
SB
53909
20268
43
Triangular
mesh
27064
12068
12068
25626
25983
Quadrangular
mesh
3915
9428
9428
6716
6716
Chapter 3
Innovating Aerodynamic Deck shape
3.0. Introduction.
Based on bluff body aerodynamic, the aerodynamic coefficients affected by deck shape.
Aerodynamic coefficients are sensitive to any small changes in deck shape. So, the deck shape
should be studied extensively before deck selection. As discussed in literature experimental or
numerical analysis give approximately the same results. However, for the final stage of design,
experimental work should be conducted for verification and safety requirements.
Deck shape determines the aerodynamic forces acting on cable bridges. Bridge deck generates
the major part of wind load acting on the bridge. As deck shape get smooth, the generated
aerodynamic forces decrease. Having a stiffening bridge girder with a smooth aerodynamic
shape is a challenge. The challenge is that the traditional bridge deck shapes cannot be
considered as an aerodynamic shape such as truss girder, slab on girder or box girder bridges.
Those traditional types have sharp edges that lead to airflow separation and increase drag forces.
During 21st century, engineers studied different solutions for developing a smoother bridge deck.
First, they designed curved box girder as Sutong Bridge. They studied and designed a separated
double box girder with adding a slot in between as Stone Cutter Bridge. Finally, they proposed a
three separate boxes with two slots in between as proposed for Messina Bridge.
This chapter discusses improving and developing bridge deck cross sections. Those proposed
shapes were modeled using CFD techniques for selecting the deck shape that generated low
aerodynamic forces and calculating aerodynamic derivative for this deck shape.
1:120 of the span length. The flutter resistance can be further enhancement by open longitudinal
slots in the rod deck. Ostenfeld et.al 1970 noted that the drag of the truss section is more than
three times of the streamlined box as shown in figure (3.1)
Figure (3. 1) Drag coefficient for a truss section and streamlined box section.
Truss girders are commonly found to be 15%-20% heavier than box girder designed for the
similar live load. Also, maintenance is difficult, and costs are considerably higher.
45
CD =0.25
CD =0.90
CD =0.34
CD =0.42
Figure (3. 4) Lift forces for the symmetric and non-symmetric airfoil.
Width over depth ratio has a significant effect on aerodynamic forces; as this ratio increases drag
forces decreases and the shape is considered to be more aerodynamic.
Based on the discussion before, six deck shapes were suggested considering aerodynamic aspect.
Each deck section has three lanes and a train way per each direction as discussed in the
following.
9.16
13.05
5.42
7.37
SECTION A-A
11.18
9.16
2.00
5.42
7.37
SECTION A-A
24.96
2.00
10.50
2.00
24.96
9.00
9.16
5.42
11.18
11.18
5.42
SECTION B-B
3.2.3. Shape 3
This section is the same as Deck 2, but the gap width was increased to be 3.0 m as shown in
figure (3.9)
48
9.00
12.18
9.15
3.00
5.42
7.37
SECTION A-A
25.96
3.00
10.50
25.96
3.00
12.18
12.18
9.16
5.42
5.42
SECTION B-B
3.2.4. Shape 4
To avoid any construction issue occurring due to elliptical deck shape, Deck 4cross section was
studied. This cross section is the traditional streamlined box section. The cross section is
modified to make full use of the depth. The train will go through the truss, and the traffic will be
on the top chord. The width over depth ratio is 5.43, Figure. (3.10).
4.00
15.00
15.00
4.50
4.00
4.50
5.50
5.50
22.50
3.2.5. Shape 5
Regarding aerodynamic forces, deck 4 was modified to reduce width over span ratio to be 7.14,
Figure (3.11)
4.00
6.00
15.00
15.00
9.00
5.50
5.50
49
6.00
4.00
3.2.6. Shape 6
Deck 6 is a modification for Deck 5; Deck 6 is modified to dissipate lift forces acting on bridge
deck as shown in Figure.1 F. Two slots were added between train and truck ways; its width is 3.0
m. Width over span ratio is 7.92, Figure. (3.12)
8.50
38.50
8.50
9.00
3.00
3.00
5.50
4.00
4.00
50
52
3.3.2. Meshing:
The domain discretized into subdomains (mesh). Mesh Height (yw) values are summarized,
Table (3.2). From this table, the ratio yw to characteristic dimension is very different even Re at
the same order.
Standard k- model: The turbulent kinetic energy, k and the rate of dissipation of turbulent
kinetic energy, are modeled. Regarding literature, Standard k- model provides a realistic
picture of the flow in case of turbulent flow in pipes and channels.
1.42
1.68
1.00
1.00
4.38
53
0.0012
0.085
1.393
Wind direction
angle
Figure (3. 20) Shows drag coefficient for studied deck shaped.
Equation (3. 1)
Equation (3. 2)
Equation (3. 3)
Regarding the discussed modeling criteria, six models were prepared, and aerodynamic
coefficients were calculated and illustrated in the chapter (4). As its necessity, it is recommended
to mention that deck shape 3 generated lower aerodynamic forces than others. Deck 3 was
selected for flutter analysis as in the following section.
54
complex form
for representing the flutter oscillation. Scanlan and Tomoko were firstly
developed a linear form as in the following:
(
)(
(
).
Equation 1
/
In this form, there are six aerodynamic derivatives. Later on, Scanlan improved those equations
that involve eight aerodynamic derivatives as follows [20]:
(
)(
(
).
Equation 2
/
Where:
L is the lift forces;
M is the pitching moment;
B is the width of the bridge;
U is the mean wind speed;
is the air density;
and
are flutter derivatives
L and M cab be written as the following:
( )
( )
( )
( )
Where: CL (t) and CM (t) are the non-dimensional lift and moment coefficient at time instance t.
The following procedure can extract the aerodynamic derivatives: (1) impose a forced harmonic
motion in vertical (h) or torsion () on the rigid cross-section and carry out Finite Volume
Method (FVM) simulations using FLUENT;
(2) apply the least square method to t stable C L (t) and CM (t) curves, and obtain amplitude CL
and CM and their shift phase angle
and
.
Flutter Derivatives for pure vertical motion
are in the following:
( )
( )
( )
( )
55
Equation (3. 4(
( )
( )
( )
Equation (3. 5(
U=V=0
W2
W4
W5
W6
V1
V3
Bridge Section
V5
V6 Rigid Boundary Layer Mesh
V4
Dynamic Mesh Region
V2
Static Mesh Region
U=V=0
56
P=0
W3
Outlet Boundary
W1
The simulation covered reduced wind speed range 2<U/fB<12 at increment 2; twelve runs was
performed.
Table (3. 4) Reduced velocity and forced motion amplitudes.
U/FB
2.00
4.00
6.00
8.00
10.00
12.00
f (Hz)
10.92
5.46
3.64
2.73
2.18
1.82
hi (mm) 30.00
42.43
51.96
60.00
67.08
73.48
i
3.00
4.24
5.20
6.00
6.71
7.35
Where:
58
CHAPTER 4
Aerodynamic Results
4.1. Introduction
This chapter divided into two parts, the first one discussed proposing an aerodynamic deck shape
and the second one discussed calculating flutter derivatives for proposed deck shape 3.
Suspension bridges are sophisticated structures that have instability problem due to its flexibility.
Different parameters affect its stability from Aerodynamic forces, statical system, and girder
stiffness. The effects of parameters on the results are discussed in the following sections
59
60
61
62
63
64
Figure (4. 12) Deck shape 3 air flow vectors (2D view)
65
Figure (4. 15) Deck shape 6 pressure distribution around for the angle of attack -10 degree.
66
Figure (4. 16) Deck shape 1 velocity distribution around for the angle of attack -10 degree.
For improving the elliptical shape Deck 1, two slots were added for improving the aerodynamic
performance. Two slot widths were studied 2 meters and 3 meter Deck 2 and Deck 3
respectively; this small change in slot width reduces aerodynamic forces. Those two slots
increase width over depth ratio; that made the deck shape smoother and helped in distributing
aerodynamic forces on the Deck 2 and 3. The aerodynamic forces were almost constant despite
the change of wind direction angle. Those slots help in reducing pressure acting on the bridge
deck, figure (000); the air flow vented through slots and reduced aerodynamic forces Deck 2and
3.
Figure (4. 17) Deck shape 3 pressure distribution around for the angle of attack -10 degree.
67
Figure (4. 19) Deck shape 4 air flow vectors (2D view)
68
69
Figure (4. 23) Deck shape 5 air flow vectors (2D view)
70
71
Figure (4. 27) Deck shape 6 air flow vectors (2D view)
72
1.20
1.00
DECK 1
Cd
0.80
DECK 2
DECK 3
0.60
DECK 4
0.40
DECK 5
0.20
-10
-8
-6
DECK 6
0.00
-2
0
-4
Sutong Bridge
2
10
Angle of attack
Figure (4. 30) drag coefficient (CD) for studied deck shapes.
1.00
0.50
DECK 1
0.00
-8
-6
-4
-2
0
-0.50
10
DECK 2
DECK 3
Cl
-10
DECK 4
-1.00
DECK 5
-1.50
DECK 6
-2.00
Sutong
-2.50
Angle of attack
Figure (4. 31) Lift coefficient (CL) for studied deck shapes.
74
0.20
CM
-10
-8
-6
-4
0.10
DECK 1
0.00
DECK 2
-2
0
-0.10
10
DECK 3
DECK 4
-0.20
DECK 5
DECK 6
-0.30
Sutong
-0.40
Angle of attack
Figure (4. 32) Moment coefficient (CM) for studied deck shapes.
Numerical Results show that the elliptical shape (Deck 1) generates a lower drag forces than
other sections. As slots width increases, drag forces decreases. Elliptical deck shapes 1, 2 and 3
generate lower drag forces than closed box deck shapes 4,5and 6; and both groups generate
lower values than Sutong Bridge.
Elliptical cross-sections 1, 2 and 3 generate lower lift forces than other studded sections. As slot
width increases, lift forces are approximately constant against the change in wind angle. As
bridge deck shape get smoother, aerodynamic forces decrease. Closed truss sections generate
approximately double lift forces than elliptical cross sections. Adding two slots for Deck 5, Deck
6 generates half lift forces of Deck 5. Elliptical deck shapes generated lower moment acting on
closed truss. For decks 4, 5 and 6 as B/d ratio increases, aerodynamic forces decreases. Results
proof that as bridge deck cross-section gets smoother aerodynamic forces generated by the deck
decrease.
75
The forced motion simulated the bridge deck motion under wind loads. The bridge oscillations
were divided into two separate motions vertical and torsional motions representing bending and
torsional motion respectively. As the bridge oscillated vertical or torsional, the velocity and
pressure distributions mutated, figures (4.33 to 4.52). This caused a change in aerodynamic
forces and calculated aerodynamic coefficients.
Figure (4. 33) Pressure distribution around deck under forced vertical motion at 0.05 second
Figure (4. 34) Pressure distribution around deck under forced vertical motion at 0.10 second
76
Figure (4. 35) Pressure distribution around deck under forced vertical motion at 0.15 second
Figure (4. 36) Pressure distribution around deck under forced vertical motion at 0.20 second
Figure (4. 37) Pressure distribution around deck under forced vertical motion at 0.25 second
77
Figure (4. 38) Pressure distribution around deck under forced vertical motion at 0.30 second
Figure (4. 39) Pressure distribution around deck under forced vertical motion at 0.35 second
Figure (4. 40) Pressure distribution around deck under forced vertical motion at 0.40 second
78
Figure (4. 41) Pressure distribution around deck under forced vertical motion at 0.45 second
Figure (4. 42) Pressure distribution around deck under forced vertical motion at 0.50 second
Figure (4. 43) Air velocity distribution around deck under forced vertical motion at 0.05 second
79
Figure (4. 44) Air velocity distribution around deck under forced vertical motion at 0.10 second
Figure (4. 45) Air velocity distribution around deck under forced vertical motion at 0.15 second
Figure (4. 46) Air velocity distribution around deck under forced vertical motion at 0.20 second
80
Figure (4. 47) Air velocity distribution around deck under forced vertical motion at 0.25 second
Figure (4. 48) Air velocity distribution around deck under forced vertical motion at 0.30 second
Figure (4. 49) Air velocity distribution around deck under forced vertical motion at 0.35 second
81
Figure (4. 50) Air velocity distribution around deck under forced vertical motion at 0.40 second
Figure (4. 51) Air velocity distribution around deck under forced vertical motion at 0.45 second
Figure (4. 52) Air velocity distribution around deck under forced vertical motion at 0.50 second
82
83
Figure (4. 58) Pressure distribution around deck under forced Torsional motion at 0.05 second
84
Figure (4. 59) Pressure distribution around deck under forced Torsional motion at 0.10 second
Figure (4. 60) Pressure distribution around deck under forced Torsional motion at 0.15 second
85
Figure (4. 61) Pressure distribution around deck under forced Torsional motion at 0.20 second
Figure (4. 62) Pressure distribution around deck under forced Torsional motion at 0.25 second
86
Figure (4. 63) Pressure distribution around deck under forced Torsional motion at 0.30 second
Figure (4. 64) Pressure distribution around deck under forced Torsional motion at 0.35 second
87
Figure (4. 65) Pressure distribution around deck under forced Torsional motion at 0.40 second
Figure (4. 66) Pressure distribution around deck under forced Torsional motion at 0.45 second
88
Figure (4. 67) Pressure distribution around deck under forced Torsional motion at 0.50
Figure (4. 68) Air velocity distribution around deck under forced Torsional motion at 0.05
second
89
Figure (4. 69) Air velocity distribution around deck under forced Torsional motion at 0.10
second
Figure (4. 70) Air velocity distribution around deck under forced Torsional motion at 0.15
second
90
Figure (4. 71) Air velocity distribution around deck under forced Torsional motion at 0.20
second
Figure (4. 72) Air velocity distribution around deck under forced Torsional motion at 0.25
second
91
Figure (4. 73) Air velocity distribution around deck under forced Torsional motion at 0.30
second
Figure (4. 74) Air velocity distribution around deck under forced Torsional motion at 0.35
second
92
Figure (4. 75) Air velocity distribution around deck under forced Torsional motion at 0.40
second
Figure (4. 76) Air velocity distribution around deck under forced Torsional motion at 0.45
second
93
Figure (4. 77) Air velocity distribution around deck under forced Torsional motion at 0.50
second
2.00
1.50
1.00
0.50
0.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
-0.50
-1.00
-1.50
CL
Forced Motion
Figure (4. 78) forced non-dimensional vertical bending (h/B =0.05) simulated and Corresponding
CL time history (U/fB =6).
94
0.80
0.60
0.40
0.20
0.00
-0.20
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
-0.40
-0.60
-0.80
CM
Forced Motion
Figure (4. 79) forced non-dimensional vertical bending (h/B =0.05) simulated and Corresponding
CM time history (U/fB =6).
6.00
4.00
2.00
0.00
-2.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
-4.00
-6.00
Cm
Forced Motion
Figure (4. 80) forced non-dimensional torsional bending 0 simulated and corresponding CM time
history (U/fB =6).
Aerodynamic derivatives obtained as equations 4 and 5 and shown in Fig. 20 and 21.
95
0
0
10
12
-2
-4
H1
H2
-6
H3
-8
H4
-10
-12
U/fB
A2
A3
A4
2
0
0
6
U/fB
10
12
Area
Bending Inertia (I)
Horizontal Inertia (I)
Torsional Inertia
(m2)
(m4)
(m4)
(J)
1.5
8.7245
2295
10
97
CHAPTER 5
Innovating ultra-long span bridge
5.0. Introduction.
Girder bridges have existed for millennia in a variety of forms depending on resources available.
The oldest types of bridges are the beam, arch and swing bridges, and they are still built today.
These types of bridges have been built by human beings since ancient times, with the initial
design being much simpler than what we enjoy today. As technology advanced the methods were
improved and were based on the utilization and manipulation of rock, stone, mortar and other
materials that would serve to be stronger and longer.
A plate girder is a girder that has been fabricated by welding plates together to create the
desired shape. The fabricator receives large plates of steel in the desired thickness, then cuts the
flanges and web from the plate in the desired length and shape. Plate girders can have a greater
height than rolled steel girders and are not limited to standardized shapes. The ability to
customize a girder to the exact load conditions allows the bridge design to be more efficient.
Plate girder can be used for spans between 10 meters and more than 100 meters (33 feet to more
than 330 feet). Stiffeners are occasionally welded between the compression flange and the web
to increase the strength of the girder.
98
99
68.42
9.00
12.18
3.00
5.42
9.15
7.37
SECTION A-A
25.96
3.00
10.50
9.16
25.96
3.00
12.18
12.18
5.42
5.42
SECTION B-B
II
33.00
0.75
0.75
0.75
Var.
Var.
0.75
II
Calculate Loads
Unsafe
Check
stresses
safe
Start
Unit weight
T
Properties
200 GPa
77 KN/m3
0.3
350 MPa
24.87 GPa
24 KN/m3
0.3
30 MPa
205 GPa
82.4
KN/m3
1.6 GPa
5.1.4. Loads
5.1.4.1. Dead Loads
Own weight was calculated from members weights; that was calculated on an iterative
process. The section was assumed then own weight was calculated, and straining action and
actual stresses were calculated. If the elements stresses were lower than allowable stress,then the
assumed sections were chosen; else, the assumed sections were increased until resultant stresses
were lower allowable stresses.
Super Imposed Dead Load (SIDL) was assumed to be 250 kg/m2.
5.1.4.2. Live loads
Codes and standards provide live load models for short and medium span bridges. However, the
models are not convenient for long span bridges due to the change of the structure type and
traffic pattern. The critical traffic does not depend on the heavy truck only. The heavy traffic
does not have significant influence on the bridge design. Moreover, some vehicles, their weight,
and percent of loaded span led to a search for predicting real traffic data. A database for traffic
data were collected from different places, and then a numerical procedure for predicting traffic
jam scenarios was developed using the database.
Observing traffic statistics in the United States for the last 30years, it was found that the number
of the vehicle miles logged annually on American highways has increased 225%, with heavy
truck traffic increasing 550%. Moreover, some percentage of trucks runs overweight, particularly
if it is to their economic advantage.
This study was implemently the Euro Code for designing deck cross section and truss elements.
The Eurocode was selected due to it is widely usage and may be implemented in the future
abroad. The Euro Code 1 Part 2 applies to bridges with spans from 5 to 200 m, and carriageway
width up to 42 m. It presents four models for determining the main vertical loads from traffic.
5.1.4.2.1. Traffic Load Models
(1) Load Model 1 (LM1) consists of concentrated and uniformly distributed loads, that cover
most of the effects of the traffic of trucks and cars, figure (5.10). The code specifies live load to
be used in each traffic lane. Therefore, there is no need to introduce multilane factors. It is used
for general and local verifications.
Location
Axle Load
Distributed load
Lane 1
300KN
9KN/m2
Lane 2
200KN
2.5KN/m2
Lane 3
100KN
2.5KN/m2
Other lanes
0
2.5KN/m2
Remaining areas
0
2.5KN/m2
103
* +
Equation (5. 1)
Figure (5. 12) Load Model 71 and characteristic values for vertical loads.
This load is not the actual burden, or the real dynamic load effect of those loads should be
considered. The dynamic factor that increases the static load effect depends on the level of
maintenance of trucks.
For carefully maintained trucks, it is
Equation (5. 3)
Where
is the determinant length associated with . This study assumed maintained trucks.
was calculated as three times cross girder spacing (Table 8.2a Eurocode1
For the elliptical deck was calculated as following:
Equation (5. 5)
Equation (5. 6)
Equation (5. 7)
Equation (5. 8)
Wind load acting on the bridge deck was divided into three loads Drag, Lift, and moment acting
in horizontal, vertically and torsionally, respectively. The wind direction angle equaled zero; for
105
zero angles of attack the horizontal forces represented the drag forces and the vertical forces
represented the lift force. The aerodynamic coefficients were 0.21; 0.1 and -0.05 for drag, lift and
moment, respectively.
For calculating the wind forces, the wind velocity was required. The designed wind speed was as
Sutong Bridge 40m/sec. Moreover, by substituting in equation 5.6,5.7, and 5.8 static
aerodynamic forces were as in table (4.1).
Table (5. 4) static aerodynamic forces acting on the bridge Deck
Drag forces (t/m)
13.37
lift forces (t/m)
6.37
Moment forces (t.m/m)
-41.4
Those forces were applied as distributed load acting on main elements of the bridge as elliptical
beams, longitudinal beams and box section. For the box section, the acting wind loads were
applied as distributed load. For determining the distributed load values, the wind load was
divided by the subjected deck length, and then each member carried loads acting on it from
centerline to centerline. For moment load, it was applied as twist moment acting on each
elliptical beam, and the value equaled the spacing multiplying by acting moment value.
106
107
The box meshed into approximately 1m2 for accuracy; the box was modeled as a shell element,
figure (5.16). Frame elements were divided into small elements to connect shell elements nodes.
This study concerns only with the preliminary design and dimensioning only.
108
G: Modulus of rigidity;
L: beam length.
Figure (5. 19) Deck model sketch for calculating torsional stiffness.
Table (5. 5) Proposed deck properties
Area (m2)
1.5
4
Bending Inertia (I) (m )
8.7245
Horizontal Inertia (I) (m4)
2295
Torsional Inertia (J)
10
110
compression. Axially loaded members are more efficient than flexural members; that contributes
to the economy of a cable-stayed bridge.
(Harp)
(Fan)
111
(Semi-fan)
Figure (5. 21) cable arrangements
A harp-type cable arrangement offers a clean and delicate appearance because an array of
parallel cables will always appear parallel irrespective of the viewing angle. It also allows an
earlier start of girder construction because the cable anchorages in the tower begin rear to the
pylon.
A fan-type cable arrangement can be very attractive, especially for a single-plane cable system.
Because of the cable slopes are steeper, the axial force in the girder, which is an accumulation of
all horizontal components of cable forces, is smaller. This feature is advantageous for longer
span bridges where compression in the girder may control the design.
The semi-fan cable stay bridge was used for different bridges all over the world due to its
efficiency, figure (5.22). In semi-fan system, the cables are distributed over the pylon
The cable-stayed bridge can be simulated as girder elastically supported (beam on elastic
foundation- springs supports). The stayed cables act as an elastic support.
Studies discussed the three systems of cable stays from different aspects as deck, cable and pylon
forces. The forces were translated into required material and costs. The cost of the deck for the
semi-fan system was the lowest, figure (5.22). Olfat Sarhang 2012 studied different parameters
for comparing those three systems and to what extent various parameters affect the cost of the
bridge. Pylon height to span ratio was one of the studied parameters, and it was found that the
optimum ratio was varied from 1 to 1.3 of the central span. The cost of the pylon increases as
this ratio increased.
Also, the number of stays cable affects the cost of the bridge, as number of stay cables increases
the cost of the bridge increases, figure (5.23)
112
Figure (5. 22) Variation of the deck cost with numbers of stay cables in three types of cablestayed bridge.
Figure (5. 23) Variation of Pylon cost with some stay cables in cable-stayed bridges.
113
required design stage; where different models may be used based on the level of required details.
For more design details and safety aspects, the full-scale model (Physical model) should be
prepared as discussed in section (2.3.2).
As mentioned in literature, for modeling the whole bridge, the frame element is an appropriate
element for modeling pylon and girder; where pylon and girder behave like a beam. However for
the more detailed design of the deck, the shell model will be appropriate as illustrated in section
(5.1). The cable also may be modeled as frame element with a very small bending inertia, and
idealized modulus of elasticity (Ernsts Modulus) as will be discussed in section (5.3.1). That
reduction in modulus of elasticity is to consider cable sag effects on cable stresses; in that
condition, the cable should be under sufficient tension under permanent loads; so that, any
compression takes place under live loads will consider as a reduction of the initial tension.
Where:
: is the cable material effective modulus;
: is the horizontal projected length of a cable;
is the weight per unit length of the cable;
is the cross-sectional area of the cable;
is the tension in the cables.
of intermediate supports, which may constrain lateral deformations. There are two ways of
increasing the lateral stiffness, the first one is modifying the deck shape and improving its
stiffness; the second is improving the statical system to increase its ability for resisting lateral
loads, this research will focus on the second one.
Suspension bridges are cable structures, which were widely used for several decades as its
construction speed and minimum use of materials, cables have only axial stiffness and carry only
tension force (Ahmadizadeh 2013). Cables are relatively lightweight structures that paid for its
use in large spans structures; moreover, cables have several advantages as high strength high
degrees of flexibility, and elastic behavior, the possibility of pre-tensioning and cost-effective
construction (Salehi Ahmad Abad et al. 2013). On the other hand, high geometric nonlinearity
has always been a potential threat to the stability of cable structures; that makes the analysis and
construction of cable-stayed systems are more challenging than most normal structures. Many
researchers paid attention to the analysis of cable structural systems(Salehi Ahmad Abad et al.
2013). Cable structures are dependent on applied loads, which cause large displacement and
rotation until reaching the equilibrium state. This research focused on cable suspension bridges.
In suspension bridges, hangers transfer loads from the deck to the main cable. The shape of the
main cable is the design key; the cable has a curved shape. For an approximation, the shape was
assumed to be a parabolic shape. Several methods discussed the analysis of suspension cables
analytically and numerically. However, those methods focused on the analysis of vertical
suspension cables and neglecting out-of- plane hangar inclination.
Hangers are commonly used as vertical hangers in traffic bridges. However, they can be used in
vertical or inclined in pedestrian bridges. The inclined hangers were commonly used in
pedestrian bridges as Harbor Drive at san-Diego, California, USA, Peace Bridge at Northern
Ireland and Gateshead Millennium Bridge at England. The inclined hangers act better than
vertical for dynamic and lateral loads, On the other hand, the excessive tensile strength leads to
fatigue failure that requires modification in their systems to achieve optimum system (Barghian
2011). The use of inclined hangers in traffic bridges is limited; the simple form of inclined
hangers in traffic bridges are the arch with inclined hangers as Network Arch Bridge.
Main cable curve profile is the analysis key for designing the suspension bridges;
The classical approach for cable analysis was deflection theory; while, as a result of the
development of finite element analysis (FEA), the profile finding method has mainly nonlinear
finite element method (Haixin HUANG et al. 2013). Several analysis approaches were developed
for numerical analysis as the parabola method, the segmental parabola method as well as the
segmental catenaries method.
J.Visontai (J.Visontai 1972), proposed an approximate analysis for suspension structures with
cable in inclined plane subjected to vertical and horizontal static loads. The proposed method
was valid for narrow suspension tube bridges under uniformly distributed static loads, and the
cable was subjected to horizontal wind load. This approximate method was based on energy
method and deflection theory. He studied vertically loaded suspension structures having inclined
cable plans with and without horizontal loads; the accuracy of the analysis was about 12%.
116
L.Greco et al., (Greco et al. 2014) proposed an approach to the static analysis of cable structures
using the form finding method; this approach carried on two steps form finding and structural
analysis. The approach was verified by numerical analysis based on the catenary element. It
relied on three-dimensional vector form solution of the catenary equation. Moreover, the
catenary element was an appropriate element for cable modeling different from the pseudo-linear
approach and modified modulus element (Ferri et al. 2006). The initial cable shape was
determined by form finding using Newton-Raphson method, and the initial stresses were
calculated by forced catenary method under self-weight and cable pre-stress. As the initial cable
shape was closer to the final shape, the iteration steps decrease; also, the iteration method affects
consumed time. Jakub et al. (Vasek & Sucharda 2013), studied calculating of parabolic catenary
using four iteration method; and compared the results of those methods. Those methods were
direct iteration, regular falsi, bisection and tangential methods. They used Mat lab software for
calculating the actual catenary length. A thirty-meter cable span was subjected to 0.8 Kn/m'
gravity directional load. They found that bisection method consumed the least time; the
tangential method used minimum iteration steps Newton method was an extensive iteration step
that calculates derivation on each point.
4.3.2.1. Suspension cable with vertical hangers
The main span was divided into two zones AC and BC, figure (4.23). For simplifying the
analysis, the following assumptions were considered. Figure (4.24) shows analysis algorithm and
flow chart
Y
VA
H
VB
H
2
3
4
5
C
a1
a1
a2
a3
a3
L/2
L/2
117
Figure (5. 27) analysis diagram for determining cable profile and forces
Assumptions:
Inextensible and weightless cable;
The cable was divided into sub-elements as straight frames between two hangers, and the
distributed load was assumed to be like an equivalent concentrated load at each node;
The girder system was simply supported girder hanged with equally spaced hangers;
The analysis started with the global stability of the cable, then the internal stability at each node
of the cable as the following:
1. Calculate vertical and horizontal reactions at supports A, B and internal tension force (T) in
the cable;
2. Start from node no. 0 at support A the initial values of the sum of forces applied to the cable;
the equilibrium applied at each node based on the initial shape or the previous iteration and the
new coordinates calculated;
3. Move to the next node, a segment of cable added and the equilibrium applied for the node
(i+1) as in step 2;
4. Repeat steps 2 and 3 for all nodes along the cable till support B;
5. Recalculate the actual length of the cable and the applied loads and repeat steps 2, 3, 4 and 5
until convergence.
The first iteration:
(
)(
)(
Where
is cable own weight;
the cable, which equals
)
(
(
)
)+
(
)
Equation (5. 30)
Where
(
)]
119
calculated as
For each hanger, i, the inclination angle calculated by the following equation, and figure (4.26)
illustrate the angle.
Equation (5. 35)
120
zi
zi
xi
xi
Suspension cable
Deck
Hanger
For small horizontal sag, the horizontal force in cable neglected; if it was calculated, it
would be very big which is illogical; it was assumed to be normal forces in the cable.
Three-dimensional suspension cable (in-extensible)
In pedestrian suspension bridges, the own weight of the cable is lower than
In suspension bridges (traffic uses), the main cable carries all vertical loads and the cable; the
own weight of the cable cannot be neglected; that made the analysis of the suspension cable is
more complicated.
For a single segment of the main cable (between two hangers), the equilibrium equations were
solved at each node of the hangers in three dimensions.
121
Wx
Wz
Where Fx and Fz are equivalent loads and M is the equivalent twist moment; for small , the
acting twist moment gets lower and may be neglected. Those forces will be in equilibrium with
hangers load. However, for large
the twisting moment caused local instability problem at
nodes due to the low torsional stiffness of the cable; that leads to cable rotation and horizontal
displacement corresponding. To overcome this displacement, transverse cables connect between
main suspension cables can be added; those transverse cables restrict the horizontal
displacement, and internal tension forces took place in those cables.
For verifying the procedure above, the following examples were studied; the proposed method
were used for estimating initial cable shape (profile) then an iteration process for tuning bridge
deck profile. Also, the effect of using inclined hangers and three-dimensional suspension cable
on the global stability of the bridge.
123
)(
( ) )
bending frequency; m is the mass per unit length, r is the radius of gyration
the bridge width and
is the air density at the bridge location.
124
and B is
The Pylon was designed to fulfill A-shape pylon; also, the pylon carried a three-dimensional
suspension cable. The two supporting systems had different inclined angles. The stayed cables
were hanged at the centerline of the pylon and suspension cable was supported at the edge of the
pylon, figure (4.34).
127
(a)
(b)
Figure (5. 38) Proposed pylon cross section a. Typical section; b. section with transverse
stiffener.
The section properties were calculated using AutoCAD software. The actual cross section were
drawn and the real properties calculated using mass properties command.
5.2.2.2. Deck.
The proposed deck discussed in Chapter 3 was used in the analysis.
5.2.2.3. Cable supporting system
The proposed cable supporting system was a hybrid system, which is a combination of the cable
system and suspension bridge as discussed in section (4.2.3.) and the following more details
about chosen suspension and stayed cable systems will be discussed. Moreover, the load path
analysis for such three-dimensional cable system was a complicated problem; so, for simplicity,
the suspension cable was responsible for the carrying all vertical loads and stayed cables were
supposed to improve bridge lateral stiffness. Also, for the safety aspect, the stayed cables were
designed for carrying half of the vertical load for each loaded span; however, the initial strain for
stayed cables were almost zero. The following sections discussed proposed cable systems (stayed
cables and suspension)
5.2.2.3. Stayed cables.
In the proposed system, stayed cables carried about 500 m beside the pylon from the central
span, figure (5.41). Although, the main aim of adding stayed cables are improving lateral
stiffness.
Regarding the pylon configuration section 4.2.2.1, the stayed cables angles were varied from 4 to
5 degrees.
129
of the central
span;
Horizontal sag: the Horizontal sag was proposed to vary from 20 and 25 m;
Inclined hangers angle: the hangers inclination angle varied depending on the horizontal and
vertical sag for the main cable at the location of the hanger. The inclination angle was calculated
for the first hanger (near the pylon); and for ensuring that hangers inclination angles will not go
far from the average, the angle was assumed to be the same at mid-span, and then the horizontal
sag was calculated backward after calculating the horizontal component of the hanger. Also, the
proposed inclination angle was calculated modified till fulfill the required horizontal sag as
summarized in Table (5.4)
Table (5. 7) studied parameters.
Model Model Model Model Model Model Model Model
1
2
3
4
5
6
7
8
Hz. Sag
25
25
25
25
20
20
20
20
Vl. Sag
300
300
300
300
300
300
300
300
cable stay
NA
NA
NA
NA
Top
NA
NA
NA
NA
transverse
The Model 1 resist more wind speed than other models, so that, the model was studied under
static loads to check the bridge serviceability. The static analysis concern different wind speeds
varied from 15 to 40 m/sec. and the deformation was compared with Euro code and AASHTO
standard; while the allowable deformation is L/600 for traffic bridges; this ratio increased to
L/800 for train bridges. Chapter 6 will represent the results.
131
CHAPTER 6
Structural analysis results
6.1. Suspension bridges.
The following table discussed the natural frequencies torsional and bending- for the studied
models in section 4.4.1.; then critical wind speed was calculated using Slebrags equation as
discussed in chapter 4.
Table (6. 1) natural frequencies and critical wind speed
Frequencies
Critical wind
(m/Sec)
fb
ft
Model 1
0.13440
0.2096
67.82
Model 2
0.1526
0.2355
75.64
Model 3
0.2157
0.325
102.5
Table (6. 2) Models Major results
Vl. Disp. Hz.
Cable Cable axial Top Beam
Hanger
Midspan Disp.
force (ton)
Axial force(ton) force (Ton)
Mid span (m)
Model 1
-0.018
0.2
20353
116.89
296.88
Model 2
-0.015
1.9
20654
21.52
321
Model 3
-0.02
0.7
24070
30
297
The proposed methods were used, and it was found that the ratio between vertical displacements
at mid span to central span were about 0.03% for studied models. The cable displacement was
insignificant that may be a result for small horizontal sag assumed; on the other hand if the
horizontal sag increased, the torsional moment acting on the main cable increased; that leads to
main cable horizontal displacement. This displacement is significant; transverse cables linked the
suspension cable may be added to overcome this displacement. The pylon shape used in
suspension models improved lateral stiffnesses of the pylon
The axial force in the main cable for model 3 increased by 19%; while the axial forces for model
2 reduced by 1.5% relatively to model 1. In the model 3, the tension forces in the top pylon beam
decreased due to that horizontal force acting perpendicular to bridge longitudinal axis derived the
two main cables from getting closer. The bridge deck cross-section was subjected to tension
force perpendicular to longitudinal axis acting on the bridge deck cross-section due to the
horizontal component of the inclined hanger.
The critical wind speed increased by 11.55 % and 51 % than vertical suspension cable and
hangers for inclined hangers and three-dimensional cable with inclined hangers respectively
132
133
134
Figure (6. 5) Suspension model bending moment diagram for the bridge deck
135
Figure (6. 6) Suspension model Shear force diagram for the bridge deck
136
Figure (6. 7) Suspension model Shear force diagram for the bridge deck
137
Figure (6. 8) Suspension model bending moment diagram for the Pylon
138
Figure (6. 9) Suspension model Shear 3-3 force diagram for the Pylon
139
Figure (6. 10) Suspension model axial force diagram for the Pylon
140
141
Figure (6. 12) Suspension model 3 Symmetrical Torsion mode shape fb 0.23792 Hz
142
Figure (6. 13) Suspension model 3 Axi-Symmetrical Torsion mode shape fb 0.2442 Hz
143
144
145
120000
100000
80000
60000
20000
0
Model Model Model Model Model Model Model Model
1
2
3
4
5
6
7
8
Figure (6. 14) forces in main cable and tower top beam
146
(a)
(b)
Figure (6. 15) deformed shape under dead loads a. Model 1 and b. Model 2.
147
(a)
(b)
Figure (6. 16) deformed shape under dead loads a. Model 3 and b. Model 4.
148
149
150
Figure (6. 19) Hybrid Bridge Axial force for the Pylon
151
Figure (6. 20) Hybrid Bridge Axial force for the main cable
152
153
154
155
156
For determining an approximate value for the critical wind speed, selbrages equation
was used as discussed in section (2.2.4.). The pure bending mode shape occurred, and
it was easy for visualization as shown in figure (6.21). However, the pure torsion
mode shape did not happen; and the first combined torsion and sway mode shape was
used for determining critical wind speed as shown in figure (6.22 and 6.23). Table
(6.6) summarizes studied models natural frequencies and the calculated critical wind
speed using Selbrages equation. The disappearance of pure tension may be due to
that the deck was under tension along the transverse direction of the bridge; those
tension forces may restrict pure torsion and led to a combination of torsion and sway.
Table (6. 6)) natural frequency (Hz) and critical wind speed (m/sec).
Mode
Mode
Critical wind speed
Ratio
Ft (Hz)
Fb (Hz)
No.
No.
(m/sec)
(Ft/Fb)
M1
106
0.28855
104
0.19074
1.512792
87.61
M2
105
0.25400
104
0.18724
1.356548
69.45
M3
202
0.26750
201
0.19714
1.356904
73.17
M4
202
0.27231
201
0.19403
1.403443
77.31
M5
103
0.27624
102
0.19534
1.41415
79.04
M6
103
0.26665
102
0.19529
1.365405
73.47
M7
202
0.26654
201
0.19814
1.34521
72.14
M8
202
0.26688
201
0.19617
1.360453
73.22
The critical wind speed for models 1 to 4 were greater than their isotopes models 5 to
8; that was a result of the increase in horizontal sag; As the horizontal sag increases,
the critical wind speed increases. In the opposite, the axial force in the main cable
increased as the horizontal sag decreased; the axial force increased by 10%. The axial
force in the pylon top beam decreased by 17% approximately, when the horizontal sag
decreased from 25 to 20 meters.
For serviceability assurance, the model M1 was studied under working loads (Dead
loads and wind loads) considering different wind speed at constant wind direction
angle 0 degree. Figure (6.25and 6.26) shows the vertical and horizontal displacement
respectively. Codes do not specify a specific limitation for long span bridges; they
only recommend L/600 for vertical deformations for traffic bridges and L/800 for
vertical deformation for train bridges. For the case studied this value will be as
follows:
1.4
1.2
1
0.8
0.6
0.4
0.2
0
0
500
1000
1500
2000
2500
3000
6
5
4
3
2
1
0
0
500
1000
1500
2000
2500
3000
Lateral Displacement
vertical Displacement
Figure (6. 27) Vertical and Horizontal displacement for Tsing Ma Bridge.
CHAPTER 7
Summary and Conclusions
7.1. Summary
This research studied the aerodynamic stability of cable-stayed suspension bridges.
The research has two major axes aerodynamic and stability analysis.
The aerodynamic analysis (Axis) focused on developing a smooth deck shape that
generates low aerodynamic forces. Six deck shapes were proposed and studied using
CFD techniques, a numerical method that was verified and recommended in the
literature; three deck shape have an elliptical shape, and three has the traditional box
girder. This research has firstly introduced elliptical deck shape. The idea of this
shape was to use an aerodynamic shape (Deck shape 1), already used as a wing; the
proposed elliptical shape was modified to fulfill traffic requirements three lanes per
each direction with two train way. The elliptical shape was modified for reducing
aerodynamic forces by adding two slots; the slot width was varied from 2 to 3
meters(Deck shape 2 and 3). The traditional box section deck shape was modified by
increasing width to depth ratio (Deck shape 4 and 5). Also, slots were added for
reducing aerodynamic forces (Deck shape 6). The results of the studied deck shapes
were compared to realistic wind tunnel analysis for Sutong Bridge, and it was found
that the elliptical shape generated low aerodynamic forces. Then, aerodynamic flutter
analysis was performed for calculating flutter derivatives based on Scanlans theory
as verified in the literature. Those flutter derivatives will be used for calculating
critical wind speed for proposed statical systems.
Structural Analysis (axis) was to improve the stability of cable supporting system. A
combination of three-dimensional suspension cable and cable-stay systems studied.
Catenary theory was improved for determining three-dimensional suspension cable
profile. Three models suspension bridge and eight models cable-stayed suspension
cable were studied. The use of three-dimensional suspension cable improved critical
wind speed. In addition, the analysis worked for devolving elliptical bridge deck
shape by combining truss and box girder systems; that development improved girder
torsional and bending stiffnesss
7.2. Conclusions
The thesis had two major axes aerodynamic and structural analysis of long span
bridge as discussed before.
From aerodynamic aspects:
The elliptical deck shape was a smooth shape; that generated lower
aerodynamic forces than traditional bridge box girder.
Adding slots to the closed elliptical shape increased aerodynamic forces little
bit; however, slots made aerodynamic forces almost constant for different
wind directions.
Adding slots decreased generated moment and lift forces. However, drag
forces increased.
Slot edges may be sharp comparing to leading or trailing edges of the deck.
As slot width increases, the generated aerodynamic forces increase.
As deck shape edges get smoother, the aerodynamic forces decrease.
For deck design analysis:
The truss-box sections combined system improved deck properties and
increased torsion stiffness of the bridge deck.
Truss elements optimization help in reducing deck own weight.
The type of truss joints affects elements cross section that has a direct effect
on deck own weight.
For overall structural stability:
The use of three-dimensional cable improved the overall stability of the
bridge.
The proposed criteria for determining three-dimensional cable profile using
superposition of catenary method was efficient for weightless cable
Moreover for considering the cable own weight, the proposed method gave an
approximate initial location for the cable profile, and then an iteration method
was used for tuning the deck profile by modifying the initial tension force in
the suspension cable.
Three-dimensional cable restricted the pure torsional mode shape; that may be
due to the tension forces in the deck section perpendicular to longitudinal axis
under dead loads.
The used of three-dimensional cable made the torsional mode away from the
bending mode shape.
In the case of using three-dimensional cable, adding top transverse elements
connected the two primary suspension cables was recommended for restricting
main cable sway.
The use of a three-dimensional cable with top transverse elements connect
main cables improved the bridge stability and made the torsional mode away
from the bending mode.
As horizontal sag increased, the critical wind speed increases; and the axial
tension force in the main cable decreases.
Conduct optimization for elliptical deck design to reduce the own weight of
the bridge.
Flutter detailed analysis for the proposed cable support system.
References
Allan. Larsen, Jens H. Walther. 1998. Discrete Vortex Simulation of Flow around
Five Generic Bridge Deck Sections. Journal of Wind Engineering and Industrial
Aerodynamics 77-78: 591602.
Bartoli, Gianni et al. 2008. Innovative Solutions for Long-Span Suspension
Bridges. Bluff Bodies Aerodynamics and Applications: 14.
Dabo. Xin, Jinping. Ou, Hui. Li. 2010. Flutter Stability Analysis of Long-Span
Bridge Based on Numerical Calculation Applying Dynamic Mesh Technique. Civil
Engineering.
Engineering, Structural et al. 2007. Flutter Analysis of Long-Span Bridges Using
ANSYS. 10(1): 6182.
http://apps.webofknowledge.com/full_record.do?product=UA&search_mode=General
Search&qid=2&SID=T2p8khc45oBFPFjMlJg&page=26&doc=251&cacheurlFromRi
ghtClick=no.
Flix Nieto, Alberto Zasso, Daniele Rocchi, and Santiago, and Hernndez. 2008. Cfd
Verification of Aerodynamic Devices Performance. In Milano, Italy: BBAA VI
International Colloquium on: Bluff Bodies Aerodynamics & Applications, 2024.
Gianni Bartoli, Piero DAsdia, Sofia Febo, Claudio Mannini, Stefano Past, Lorenzo
Procino. 2008. Innovative Solutions for Long-Span Suspension Bridges. Bluff
Bodies Aerodynamics and Applications: 14.
H.Ostenfeld, Klaus, and Allan Larsen. 1992. Bridge Enginering and Aerodynamic.
In Aerodynamics of Large Bridges, , 322.
Huang, Lin, and Haili Liao. 2011. Identification of Flutter Derivatives of Bridge
Deck under Multi-Frequency Vibration. Engineering Applications of Computational
Fluid Mechanics 5(1): 1625.
Huang, Lin, Haili Liao, Bin Wang, and Yongle Li. 2009. Numerical Simulation for
Aerodynamic Derivatives of Bridge Deck. Simulation Modelling Practice and
Theory 17(4): 71929. http://dx.doi.org/10.1016/j.simpat.2008.12.004.
Jos . Jurado, Santiago Hernndez, Flix Nieto, and Manuel A. Mosquera. 2011.
Bridge Aeroelasticity. Sensitivity Analysis and Optimal Design.
M Keerthana, K P Jaya, S Selvi Rajan, Thampi Hephzibah , R Ravi Sankar. 2011.
Numerical Studies on Evaluation of Aerodynamic Force Coefficients of CableStayed Bridge Deck. 8(2): 1929.
Kuroda, Shinichi. 1997. Numerical Simulation of Flow around a Box Girder of a
Long-Span Suspension Bridge. Journal of Wind Engineering and Industrial
Aerodynamics 67-68: 23952.
0333 0033 0333
.
( )402
( )6 0
1
2
.
6 (0
3
0 )
(
).
4
5
.
6
.
7
-
-
2016
-
-
2016
:
/ -
-
/ -
-
/ -
-
/ -
-
/ -
-
-
-
2016