Global Analysis of the SuTong Cable-Stayed Bridge

Jiawu MIAO Rucheng XIAO Minshan PEI

PhD student, Engineer Professor Senior Engineer
Tongji University Tongji University HPDI Company
Shanghai, CHINA Shanghai, CHINA Beijing, CHINA

Xigang ZHANG Martin PIRCHER Dorian JANJIC

Chief Engineer Research Associate Director
HPDI Company University of Western TDV Company
Beijing, CHINA Sydney, Australia Graz, AUSTRIA

Summary
The SuTong cable-stayed bridge is the Primary Fairway Bridge of the Suzhou-Nantong Yangtze
River Bridge Project. It will be the most important project with a significant target to minish the
economy gap and promote balanced development between Suzhou and Nantong city in China. The
total length of the cable-stayed portion of this project is 2088 m with a 1088 m main span. The
height of the pylons is about 300 m. When completed it will be the longest cable-stayed bridge in
the world. This paper gives a brief description of the bridge and some considerations on structural
design and static analysis.

Keywords:cable-stayed bridge; steel box girder; long span; nonlinearity; damper; structural system;
global analysis.

1. Introduction
The Suzhou-Nantong Yangtze River Bridge Project is located in the southeast of the Jiangsu
Province in China. In recent years, the economy of the South part of Jiangsu Province developed
very rapidly, but the north part kept almost as before because of the obstruction of the Yangtze River.
Take the cities of Suzhou and Nantong as an example, the ratio of the GDP value per person of two
cities was about 2.65:1 in the year of 1998. Therefore, the construction of this project will be an
important link between the cities of Suzhou and Nantong with a significant target to minish the
economy gap and promote balanced development between the north and the south part of Jiangsu
Province. In this sense, this project is not only subjected to the national communication demands
but also one of pivotal political levers to eliminate local poverty and accelerate conjunct richness.
The total length of the bridge portion in this link is about 8.2 km. At the bridge site there are two
navigation channels, the Primary Fairway and the Special Fairway for the exclusive use by port
Nantong. The proposed bridge project consists accordingly of the Primary Fairway Bridge, the
Special Fairway Bridge and the approach spans. The Primary Fairway Bridge (hereinafter named
the bridge in brief) is a cable-stayed bridge. The Special Fairway Bridge is a pre-stressed concrete
continuous rigid frame bridge with a span arrangement of (140+268+140) m. And the approach
spans are some pre-stressed concrete continuous girder bridges with 75 m, 50 m and 30 m in span
lengths.
The design tasks of the project are carried out by China Highway Planning and Design Institute
(HPDI) Consultants, Inc. in cooperation with Jiangsu Provincial Communication Planning &
Design Institute, and the Architectural Design & Research Institute of Tongji University. The review
of design documents is performed by COWI Consultants and CHODAI CO., LTD. separately. The
detailed design was completed in July 2004. Construction of the project started in October 2002 and
the scheduled completion will be December 2008 or early.
2. Description of the Bridge
2.1 Span Arrangements
Taking full consideration of various geotechnical conditions at the bridge site, technical feasibility,
and especially construction ability, the selected design for the Primary Fairway Bridge is a double-
plane twin-pylon cable-stayed bridge with a continuous span arrangement of
(100+100+300+1088+300+100+100) m, as shown in Figure 1. Two auxiliary piers and one
transitional pier are erected in each side span. The main span of the bridge is 1088 m, which is the
longest main cable-stayed bridge span at present.
2088
100 100 300 1088 300 100 100
8 11@12 22@16 8 8 33@16 16 33@16 8 8 22@16 11@12 8

Nantong Suzhou

220 891 220

SF PF SF

P1 P2 P3 P4 P5 P6 P7 P8

Fig. 1 Span Arrangement (unit: m)

2.2 Girder
The bridge girder is a streamlined closed flat steel box girder. The total width including wind fairing
is 41.0 m accommodating dual 8 traffic lanes. The cross-section height is 4.0 m. The steel box is
generally stiffened in the longitudinal direction with closed steel troughs. Transverse plate
diaphragms are provided with a typical distance of 4.0 m and with smaller distances down to 2.27m
locally around the two pylons. The characteristic yield strengths of the structural steel are 345 MPa
and 370 MPa. The standard cross-section of the girder is illustrated in Figure 2. The thickness of the
skirts and stiffeners vary along the longitudinal direction of the bridge.
4.0
2
3
.
0

9.0 9.0
41.0

Fig. 2 Cross Section of the Girder (unit: m)

2.3 Pylons
The inverted Y-shaped pylons are about 300 m in height and are made of concrete grade C50
according to the Chinese standard JTJ01—89. The stay-cables are anchored inside steel boxes fixed
to the concrete by shear studs at the pylon top. The maximum segment of a cable anchorage steel
box weighs about 36 tons. The tie-beam between the pylon legs is fully post-tensioned for outward
thrust from the pylon legs under service loads and seismic load.
According to the project specifications and the review comments from COWI the cracking width of
the concrete pylon wall is controlled within 0.2 mm.
2.4 Stay Cables
The stay cables are arranged in double inclined cable planes with standard spacing of 16 m in the
central span and 12 m near the ends of the back spans along the girder. To reduce the effect of wind
loads, the cable stay systems are made of the parallel wire strand consisting of 7 mm wires, each
with a cross sectional area of 38.48 mm2. The nominal tensile strength of the cables is 1770 MPa.
Cable sizes range from a minimum of PES7-139 for the main span stays near the pylons to a
maximum of PES7-313 for the longest backstay. The longest cable is about 577 m with a weight of
59 tons.
The problem of cable vibration due to wind and possibly in combination with rain or parametric
excitation has been studied during the design process. Different measures to minimize stay cable
vibrations have been investigated including two kinds of cable surface treatments to prevent
formation of rainwater flows on the cables and internal or additional external damping devices. The
final measures will be chosen after detailed testing.

2.5 Foundations
The piers or pylon from P1 to P8 are supported by bored friction piles with diameters from 2.8 m
near the pile-head to 2.5 m away from the top along the piles. Each of P1~P2 and P7~P8 has 19
piles. 36 piles are driven for P3 and P6 separately. Each pylon of P4 and P5 is supported by 131
piles. The pile lengths vary between 108 to 116 m.

2.6 Connection between Girder and Pylons

The selected permanent connection between the girder and the pylons is accomplished by nonlinear
dampers as used for the Great Belt East Bridge in Denmark. These dampers do not confine the
displacement of the steel girder induced by temperature, moderate wind, and vehicle traffic, but
transfer the loads from the girder induced by gust, earthquake and other forces from specific load
combinations to an alternative pylon.
The dynamic characteristics of one damper is given as F
K=100 MN/m
F = C ⋅V α (1)
V is the relative velocity between pylon and girder
α is a constant equal to 0.4
C is a constant equal to 3750 kN/(m/s)0.4
δ
-δ0 0 δ0 =750 mm
A total of 4 dampers are placed at each pylon. A relative
movement of 750 mm between girder and pylon is allowed
before the displacement restriction equipments are activated.
Each of the four dampers at one pylon has a linear stiffness
of 100 MN/m in the event that a relative movement beyond
750 mm occurs. The static force-displacement relationship Fig. 3 Static force-displacement
for each damper unit is shown in Figure 3. relationship for each damper

3. Global Static analysis

3.1 Global Analytical Model
The RM2000 program [1] developed by TDV, Austria, has been used for the global analysis of the
SuTong cable-stayed bridge in the detailed design. Two other programs named QJX and BAP have
also been used for checking by the designer. The finite element model of the bridge is illustrated in
Figure 4.The structural modelling keeps accordance with the planned construction schemes. Each of
the stay-cables was divided into 8 sub-elements to consider cable-sag effects rather than approxima-
ting this effect by using effective module of elasticity. Other interacting non-linear effects such as
P-delta effect, large displacements and shear displacements were also considered in the calculation.
Creep and shrinkage effects were calculated according to the CEB/FIP 90 code. The flexibility of
the pylon foundations was modelled with spring elements. The connection between the girder and
both pylons were treated as nonlinear static spring elements with a gap value of 750 mm and a
linear stiffness of 100 MN/m.

Fig. 4 Finite element model of the bridge

3.2 Definition of the Bridge Final State and Analytical Results

An obvious characteristic of cable-stayed bridges is the ideal internal force distribution in the
completed structure which can be achieved by cable tension adjustment. Generally, the ideal final
state should be close to some basic conditions. That is to say, the strategy of cable tension
adjustment should aim to eliminate/minimise bending moments in the deck and pylons under
permanent loads. The cable force distributions should avoid dramatic variations between adjacent
cables.
For the SuTong Bridge the contribution ratio of traffic loading is heavy for deck stress and
counterweight arrangements in the back spans. The definition of the final state gives attention to
both situations with and without traffic. The bending moment envelops in the deck under dead load
and load combination 1 is shown in Figure 5. The maximum moment of the pylons (in the bottom)
at 28.0 MN-m is very small. The results illustrate the suitability of the achieved final state of the
bridge.

MZmin -138.0MNm

(1) D+Lmin

(2) D

(3) D+Lmax

MZmax 140.1MNm

Fig. 5 Bending moment envelop in deck

3.3 Stage analysis
The forward analysis with the ADDCON method implemented in RM2000 [4] was employed for all
erection stages to achieve the final situation as described above according to the construction
schedules of the designer. All temporary supports, tie-downs, and movements of derricks for
construction, temporary loading, and permanent loading were included in the model at various
stages. The equivalent static wind actions from different directions were also investigated at the
most detrimental construction stages such as the maximum double cantilever, the maximum single
cantilever and the completed bridge. The pre-camber calculation of all construction stages was
computed by RM2000 automatically. The third-order effect of pre-camber shapes apart from the
design elevation of the deck was also considered approximately.
The results of the construction stage analysis also showed that the global stiffness of the bridge is
very small before closure. For instance, the first tension of the longest stay cable in mid-span leads
to a vertical deflection of 1.3 m at the end of the cantilever. Even after closure, the superimposed
dead load (including paving, baluster, etc.) can still result in a vertical deflection of 1.8 m in the
centre of mid-span. In addition, geometric nonlinearity in general has remarkable effects, especially
for erection geometry of the deck.

3.4 Structural System and Parametrical Analyses

As mentioned before the dampers do not confine the displacement of the steel girder induced by
temperature, moderate wind, and vehicle traffic, but transfer the loads of the girder induced by gust,
earthquake, and other forces to the tower. Therefore, definition of appropriate characteristic design
parameters of the dampers including gap value, elastic stiffness, and dynamic characteristics is
critical to achieve those intentions. Relevant parametrical analyses were carried out for some
dominate load cases including static loadings and dynamic inputs.
The dynamic parameters (see 2.6) are based on the results of time history analyses for some typical
seismic inputs. The maximum relative displacement between the girder and the pylons is smaller
than 750 mm.
For static actions, a proper gap value is the governing parameter. Taking into account all of the
responses of the above loads and considering the current product specifications of large expansion
joints, a gap of 750 mm was selected to fit the design requirements. Consequently, some
parametrical analyses were carried out to define a reasonable spring stiffness according to the
relationship curve between the bending moment response at pylon bottom (the longitudinal
displacement at the end of the girder) and longitudinal wind inputs. Some results of parametrical
studies are shown in Figure 6~8 with the spring stiffness in the range of 1~1000MN/m. These
graphs illustrate the fact that a spring stiffness of 100MN/m is proper and practicable. Here an
assumption is introduced that only the dampers at one pylon of P4 or P5 may be engaged before the
other pylon. This assumption reflects construction inaccuracy and bridge deck elongation due to
temperature.
Maximum Moment at pylon bottom

1.8 MN
2500
1.7 14
Displacement at pylon top (m)

2400 1.6
12
Total reaction force

2300 1.5
of four dampers
(MNm)

1.4 10
2200
1.3 8
2100 1.2
6
1.1
2000
1.0 4
1900
0.9
2
0 200 400 600 800 1000 0 200 400 600 800 1000 0 200 400 600 800 1000
Spring Stiffness of a damper (MN/m) Spring Stiffness of a damper (MN/m) Spring Stiffness of a damper (MN/m)

Fig. 6 Maximum moment at Fig. 7 Displacement at Fig. 8 Total reaction force of

pylon bottom pylon top four dampers
For the SuTong Bridge, the connection between girder and pylons is essential to the safety of the
pylons under extreme wind and seismic loads. Therefore, based on the performed detailed
parametrical studies and some further considerations for installation tolerances and safety margins a
maximum force for one damper of approximately 10 MN (under ULS state) is assumed as one of
design requirements. Meanwhile, the comparison between the results of designer and COWI
Consultants confirmed that the material non-linearity of the pylons plays a significant role for the
resultant reaction forces under ULS state.

3.5 Geometric nonlinearity effects

The designer has given much attention to geometric nonlinearities all the way from the preliminary
design phases to the detailed design. A special study on nonlinear effects was carried out. Two
notable remarks on geometric nonlinearities are abstracted as follows.
y For the SuTong Bridge, compared with linear analysis, geometrically non-linear effects on the
completed bridge may induce an offset of 10~20 percent of the maximum / minimum stress of
the girder and the pylons under traffic loads, and the critical location can be also shifted.
y Generally, the FE model of the stay-cables employs a straight truss element with the effective
modulus of elasticity, or dividing each stay cable into many sub-elements like RM2000, or by
 the newly catenary cable element. Different means to deal with cable sag effect make a certain
effect on the fabrication of construction process. The means of the equivalent truss element
may induce a maximum offset of 0.538 m apart from the desired location at the end of stage
analyses finally, as the mentioned in reference [2]. One of the reasons is possibly inaccurate
chord force vectors in long stay cables. To cable-stayed bridges beyond 1000 m, this
simplification should be restricted, especially for erection process. Certainly, the means of
catenary cable elements is better than dividing sub-elements but within the tolerance range.
4. Concluding Remarks
The SuTong Cable-Stayed Bridge is the most critical part of the Suzhou-Nantong Yangtze River
Bridge Project. It costs about 380 million dollars. It will supply much convenience to the people
beside the Yangtze River and accelerate economy development to make the adjacent cities rich
together. By the way, the design and construction of this bridge itself also provide a very good
chance to promote cooperation and intercommunion among some famous designers at home and
abroad.
References
[1] TDV, 2003, RM-Spaceframe Users Manual, rev 8.58-02.
[2] LIANG P., Geometrical Nonlinearity and Random Simulation of Super Long Span Cable-
Stayed Bridges. A dissertation for the degree of Doctor of Philosophy of Tongji University.
Aug, 2004.
[3] MIAO Jiawu, Comparison Study on Geometrical Nonlinearity Effects for the SuTong Cable-
stayed Bridge. 2003 Conference proceedings of Bridge and Structure subdivision of Chinese
Highway Association. Sep, 2003.
[4] Janjic D., Pircher M., Pircher H. Optimisation of Cable Tensioning in Cable-Stayed Bridges,
Journal of Bridge Engineering, ASCE, v8, n3, pp 131-137, 2003