Bridging Capacity and Aerodynamic Challenge of

Long-Span Suspension Bridges

YongXin YANG Yao Jun GE

Associate Professor Professor
State Key Lab for Disaster State Key Lab for Disaster
Reduction in Civil Eng. Reduction in Civil Eng.
Tongji University Tongji University
Shanghai, 200092, China Shanghai, 200092, China
yang_y_x@.tongji.edu.cn yaojunge@tongji.edu.cn

YongXin Yang, born 1974, Mr. Ge, born 1958, and received
received his Ph. D degree from his PhD degree in bridge
Tongji University in 2002. He engineering from Tongji
has worked in Tongji University University. He is currently the
for 9 years. Lab Director and the Chairman
of Department of Bridge Eng..

Summary
As a human dream and an engineering challenge, the structural engineering of bridging larger
obstacles has entered into a new era of crossing wide rivers and sea straits, for example, Messina
Strait in Italy, Qiongzhou Strait in China, Tsugaru Strait in Japan, Gibraltar Strait linking European
and African Continents, and so on. One of the most interesting challenges has been identified as
bridging capacity limit of suspension bridges. Traditionally, bridging capacity is referring to
longitudinal single bridge span, and the feasibility study shows that either a widely slotted deck or a
narrowly slotted deck with vertical and horizontal stabilizers could provide a 5,000m span-length
suspension bridge with high enough critical wind speed. With an innovative way to improve
bridging capacity, the concept of continuous multiple main span scheme for crossing longer water
body is proposed, in particular double main span suspension bridges, including Taizhou Yangtze
River Bridge and Maanshan Yangtze River Bridge, both with two main spans of 1080m.

Keywords: Bridging capacity, aerodynamic challenge, suspension bridge, flutter stabilization,

span length limit, double main span.

1. Introduction
Although ancient suspension bridges were built in iron chains or cables in China long before the
history of steel application in the 19th century, the construction of modern steel cable suspension
bridges around the world has experienced a considerable development for more than a century. It
took about 48 years for the span length of suspension bridges to grow from 483m of Brooklyn
Bridge in 1883 to over 1,000m in George Washington Bridge of 1,067m in 1931, and had an
increase by a great factor of about 2.2. The further increase was only by a factor of 1.3 in the next
50 years from George Washington Bridge to Golden Gate Bridge of 1280m and to Humber Bridge
of 1410m in 1981, but another factor of about 1.4 was realized in Akashi Kaikyo Bridge with a
1,991m main span within 17 years in 1998. Among the ten longest span suspension bridges
completed in the world listed in Table 1, the top four and the 7th longest bridges were suffered in
aerodynamic problems in either flutter or vortex shedding, and some control measures have been
adopted to improve aerodynamic performance, for example, both slot and stabilizer in Akashi
Kaikyo Bridge, twin-box deck in Zhejiang Xihoumen Bridge, guide vane in Great Belt Bridge,
central stabilizer in Jiangsu Runyang Bridge, central slot in Tsingma Bridge [1].
With the experience gained from ten longest suspension bridges in the world, the intrinsic limit of
span length due to aerodynamic stability is about 1,500m for a traditional suspension bridge with
either a streamlined box deck or a ventilative truss girder. Beyond or even approaching this limit,
designers should be prepared to improve aerodynamic stability of a bridge by adopting some
countermeasures for girder. A widely slotted deck and a narrowly slotted deck with vertical and
horizontal stabilizers have been attempted for a 5,000m span-length suspension bridge, and two
double main span suspension bridges under construction have been discussed with the emphasis of
aerodynamic instability.
Table 1: Ten longest span suspension bridges in the world
Span Main Girder Wind-Induced Control Year
Bridge Name Country
Order Span Type Problem Measure Built
1 Akashi Kaikyo 1991m Truss Flutter Slot/Stabilizer Japan 1998
2 Xihoumen 1650m Box Flutter Twin box China 2009
3 Great Belt 1624m Box Vortex Guide vane Denmark 1998
4 Runyang 1490m Box Flutter Stabilizer China 2005
5 Humber 1410m Box None None U.K. 1981
6 Jiangyin 1385m Box None None China 1999
7 Tsing Ma 1377m Box Flutter Slot H.K. China 1997
8 Verrazano 1298m Truss None None U.S.A. 1964
9 Golden Gate 1280m Truss None None U.S.A. 1937
10 Hubei Yangluo 1280m Box None None China 2007

2. Countermeasures for Aerodynamic Instability

In general, countermeasures for aerodynamic stability can be grouped into three types, including
structural modification, aerodynamic means and mechanical device, which are briefly introduced in
the following.
2.1 Structural Modification Methods
As already mentioned in the previous section, the traditional suspension bridges with single box
girders seem to stop at the span limitation of 1,500 m. One of the most important reasons is the
higher trend of reduction of torsional stiffness with the increase of span length, which consequently
leads to a decrease of fundamental torsional frequency. In order to increase torsional stiffness of
suspension bridges, some structural modification methods have been proposed but not realized yet,
in particular cable system modifications, which can be summarized into three kinds as follows.
(a) Crossed hanger system shown in Fig. 1a [2] or combination of vertical and horizontal cross
stays in Fig. 1b [3]
(b) Mono cable system shown in Fig. 2a [2] or spatial cable systems in Fig. 2b [4]
(c) Three cable system shown in Fig. 3a or four cable system in Fig. 3b [2]

(a) Crossed hanger system (b) Vertical and horizontal cross stays
Fig. 1: Crossed stay systems
(a) Mono cable system (b) Spatial cable system
Fig. 2: Mono or spatial cable systems

(a) Three cable system (b) Four cable system

Fig. 3: Multiple cable systems
2.2 Aerodynamic Preventive Means
A parallel approach to the aerodynamic stability problem is an attempt to reduce the aerodynamic
forces based on the configuration improvement of cross sections of bridge decks, which is usually
called as aerodynamic preventive means. There are two kinds of aerodynamic means having been
used in practice, including central stabilizer and slotted deck or twin-box deck. For example,
Runyang Bridge with a main span of 1490m has adopted a 1m high central stabilizer mounted on
the deck, shown in Fig. 4 [5], in order to raise critical flutter speed from 50m/s to 55m/s, and twin
box deck was originally adopted in Xihoumen Bridge, shown in Fig. 5 [6], for the purpose of the
increase of critical flutter speed over 78m/s.

Fig. 4: Central stabilizer of Runyang Bridge Fig. 5: Twin-box deck of Xihoumen Bridge
2.3 Mechanical Devices
Mechanical devices have been proposed as possible methods for raising total amount of damping, in
particular aerodynamic damping, and can be divided into to passive control and active control
methods [7][8]. Most of passive aerodynamic dampers consist in wing profiles fixed at the section
leading or trailing edge shown in Fig. 6 to add torsional and vertical damping, as well as the cross
terms. Active control devices have never been applied in real structures, but already considered by
several researches in a feasibility stage shown in Fig. 7 [9][10]. It should be noted that, however,
before the solution is applied in a real project, the engineering feasibility must be taken into account
in advance, even though very attractive results are realized.

Fig. 6: Passive aerodynamic damper Fig. 7: Active control device

3. Super Long Suspension Bridge

Bridging capacity challenge has been identified as bridge span length limitation, in particular the
span limits of suspension bridges as a bridge type with potential longest span. The dominant
concerns of super long-span bridges to bridge designers are basically technological feasibility and
aerodynamic considerations.
3.1 General Arrangement
With the emphasis on aerodynamic stabilization for longer span length, a typical three-span
suspension bridge with a 5,000m central span and two 1,600m side spans is considered as the
limitation of span length as shown in Fig. 8. In order to improve aerodynamic stability limit, two
kinds of generic deck sections were investigated, including widely slotted deck (WS) with four
main cables in Fig. 9a and narrowly slotted deck with vertical and horizontal stabilizers (NS) in Fig.
9b. The NS cross section has a total deck width of 50 m for the 5,000 m spanned suspension bridge
while the WS provides a wider deck solution of 80 m. An outline of the dynamic analysis procedure
and its application to four sag-span ratios combined with the two above-mentioned deck
configurations is given below [11].
f

1600 5000 1600

Fig. 8: Span arrangement (Unit: m)

(a) WS Cross section (b) NS Cross section

Fig. 9: Twin-box deck sections (Unit: m)

3.2 Dynamic Characteristics

Having performed a dynamic finite-element analysis based on the structural parameters listed in
Table 2, the fundamental natural frequencies of the structures have been calculated for all four
ratios n of cable sag to span and the two deck configurations in Table 3. The fundamental vertical
bending frequencies are not influenced significantly by both deck configurations and the sag-span
ratios. The fundamental torsional frequencies vary differently with the ratio n in the two deck
configurations, in which the frequency values go up in the WS section and go down in the NS
section with the decrease of the ratio n, but it is interesting to see that the frequency ratio of torsion
to vertical bending monotonically decreases with reduction of the ratio n [1].
Table 2: Parameters of stiffness and mass and mass moment inertia
Main Cables Stiffening Girder
Section
EA (Nm2) m (kg/m) Im (kgm2/m) EIy (Nm2) GId (Nm2) m (kg/m) Im (kgm)
WS 0.61~1.12×106 2.62~4.82×104 2.36~4.33×107 4.7×1011 2.8×1011 24000 2.16×107
NS 0.61~1.12×106 2.62~4.82×104 1.27~2.33×107 8.1×1011 4.1×10 11
24000 5.40×106

3.3 Critical Flutter Speed

With the dynamic characteristics given above and the numerically identified flutter derivatives
shown in Fig. 10, the critical wind speeds of the suspension bridges were calculated by multi-mode
flutter analysis assuming a structural damping ratio of 0.5%. The results of critical wind speeds
together with the generalized mass and mass moment of inertia are summarized in Table 3. For both
deck sections the critical wind speed increases with decrease of the ratio n, although the frequency
ratio of torsion to vertical bending slightly decreases. The most important reason is the considerable
increase of the generalized properties in the aerodynamic stability analysis. The minimum critical
wind speeds for the WS and NS sections are 82.9 m/s and 74.7 m/s, respectively, whose figures can
meet the aerodynamic requirement from most typhoon-prone areas in the world [1][11].
0. 6 0. 5
0. 5 A1
0
0. 4 A2
- 0. 5
0. 3 A3
0. 2 -1
A4
0. 1 - 1. 5
0 -2
- 0. 1 - 2. 5
- 0. 2 H1
-3
- 0. 3 H2
- 3. 5
- 0. 4 H3
- 0. 5 -4
H4
- 0. 6 - 4. 5
0 1 2 3 4 5 6 7 8 9 10 11 12 13 0 1 2 3 4 5 6 7 8 9 10 11 12 13
Vr Vr
(a) Ai of WS section (b) Hi of WS section

0. 6
A1 0
0. 5
0. 4 A2 -1
0. 3 A3
-2
0. 2 A4
0. 1 -3
0 -4
- 0. 1
- 0. 2 -5 H1
- 0. 3 -6 H2
- 0. 4 H3
-7
- 0. 5 H4
- 0. 6 -8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 0 1 2 3 4 5 6 7 8 9 10 11 12 13
Vr Vr
(c) Ai of NS section (d) Hi of NS section
Fig. 9: Flutter derivatives for both sections
Table 3: Fundamental frequencies and critical flutter speeds

Ratio m (×104kg/m) Im (×107kgm2/m) fh (Hz) fα (Hz) Ucr (m/s)

WS NS WS NS WS NS WS NS WS NS
n = 1/8 6.01 6.79 5.28 2.37 0.0596 0.0594 0.0709 0.0907 82.9 74.7
n = 1/9 6.27 7.43 5.36 3.22 0.0616 0.0612 0.0721 0.0893 88.8 77.4
n = 1/10 6.73 8.33 5.92 3.29 0.0626 0.0620 0.0727 0.0865 90.9 78.9
n = 1/11 7.66 9.52 6.77 3.62 0.0626 0.0622 0.0727 0.0840 98.9 82.7

4. Double Main-Span Suspension Bridges

Traditionally, bridging capacity is referring to a single bridge span, for example, the latest record is
Akashi Kaikyo Bridge with a central span of 1991m. The tendency of bridging capacity
development is not only invention in making span longer, but also a further or new development of
the concept to make continuous multiple main span schemes.
4.1 Development of Double Main-Span Suspension Bridges
Although the concept of multi-span suspension bridge can be traced back to Oakland Bay Bridge
built in San Francisco in 1930’s, there have been only two multiple main-span suspension bridges
completed and two design proposals available in various sources. One of the completed projects is
Konaruto Bridge built in Japan in 1961, with two main spans of 160 m, and the other is Save River
Bridge built by Portuguese in Mozambique in 1965, having the span arrangement of
110+3×210+110m. The COWI firstly proposed a double-main-span suspension bridge scheme with
main spans over 1000m (1055+1100m) for Chacao Bridge in Chile in 2000, and another double-
main-span (2×1200m) suspension bridge proposal was made in the feasibility study of Qingdao Bay
Bridge in China in 2005 [12].
Under the rapid development of suspension bridges, China launched two long-span suspension
bridges with double main spans longer than 1,000m in 2007. One of them is Jiangsu Taizhou Bridge
across Yangtze River with the span arrangement of 390 + 2×1080 + 390m shown in Fig. 10, and the
other is Anhui Maanshan Bridge over Yangtze River spanned as 360 + 2×1080 + 360m in Fig. 11,
both of which are currently under construction of pylons and are scheduled to be completed by
2012 [12].

Fig. 10: Taizhou Bridge Design Scheme Fig. 11: Maanshan Bridge Design Scheme

4.2 Bridge Schemes for Comparison

In conceptual design point of views, a double main-span suspension bridge is an innovative single
main-span structure with an additional supporting pylon at the mid span to improve static and
dynamic structural performance, for example, in main cables and anchor blocks. With this central
pylon as a vertical support to main cables and stiffening girder, however, some structural
performance of this new type bridge should be improved better than that of a traditional structure.
Otherwise double-main-span suspension bridges should be avoided in practice.
In order to make the conceptual comparison of a double main-span suspension bridge and the
corresponding single main-span structure, Taizhou Bridge and Maanshan Bridge have been taking
as typical models of double main-span suspension bridges shown in Figs. 12 and 13, and the
corresponding comparison model is a traditional three-span suspension bridge with the spans of
720+2160+720m in Fig. 14. With the same steel box deck and the sag ratio of 1/9, these three
bridge models have been compared with finite element method in static and dynamic characteristics
[13].
390 1080 1080 390
+191. 500 +191. 500 +191. 500

120

120
+26. 5 +71. 500 +71. 500
+26. 5

+2. 0 +5. 92 最最最最最最 +6. 0 最最净通760× 50m +2. 0

Fig. 12: Taizhou Bridge

360 1080 1080 360
+176.850m +176.850m +176.850m

+56.850m +56.850m

Fig. 13: Maanshan Bridge

720 1080 1080 720
+296.850m +296.850m

+56.850m

Fig. 14: Single main-span suspension bridge

4.3 Structural Performance

The finite-element idealization of these five schemes was attempted with finite beam elements for
the stiffening girder and the pylons and cable elements considering geometric stiffness for the main
cables and hangers, and the static analysis was conducted with the main results shown and
compared in Table 4 under the most critical combination, dead load plus live load applied in one
main span. It can be seen from Table 4 that the cable force in double main-span is only about half of
that in single main-span, and the longitudinal displacement of the top of central pylon is quite large
compared to single main-span while the deck displacements keep in the same order between single
and double main-span structures [13].
Table 4: Structural static characteristics
Pylon Displacement (m) Deck Displacement (m) Cable Force (MN)
Bridge Scheme
Side Centre Maximum Minimum Maximum Minimum
Taizhou Bridge 0.194 1.876 4.352 3.200 184 171
Maanshan Bridge 0.189 1.790 4.214 3.142 185 172
Single main span 0.359 - 3.932 2.962 384 -

Having performed a dynamic finite-element analysis, the first and second natural frequencies of
three bridge schemes have been extracted and compared in Table. The first and second frequencies
of lateral bending and torsional vibration modes in the double main-span bridge schemes are
tremendously enhanced comparing with the single main-span bridge scheme while vertical bending
frequencies have almost the same values in both single and double main-span bridge schemes. This
is the most important reason why a double main-span suspension bridge is better in structural
dynamic characteristics than the corresponding single main-span structure [13].
Table 5: Structural dynamic characteristics
Lateral Bending (Hz) Vertical Bending (Hz) Torsion Vibration (Hz)
Bridge Scheme
First Second First Second First Second
Taizhou Bridge 0.0723 0.0971 0.0799 0.1026 0.2732 0.3598
Maanshan Bridge 0.0911 0.0934 0.0788 0.1136 0.2624 0.3457
Single main span 0.0363 0.0765 0.0960 0.1283 0.1788 0.1998

4.4 Aerodynamic Flutter Instability

For double-main span suspension bridges, one of the most challenging problems is aerodynamic
flutter instability, which takes place when a bridge is exposed to wind speeds above a certain critical
value. These critical values or critical flutter speeds for the two double-main-span suspension
bridges were predicted through the wind tunnel tests using sectional rigid models and full
aeroelastic models, and the corresponding experimental results were checked with the required
flutter speeds of 56.6m/s for Maanshan Bridge and 57.4m/s for Taizhou Bridge due to the Chinese
code. Under different angles of attack required in the Chinese code, including −3º, 0º and +3º, the
sectional model wind tunnel experiments were carried out in the TJ-1 Boundary Layer Wind Tunnel
with the working section of the 1.8 m width, the 1.8 m height and the 15 m length. The full
aeroelastic model testing was respectively conducted under smooth flow at the attack angles of −3°
and +3° with the 0° yaw angle and at the yaw angles of 0°, 5°, 10°, 15°and 20° with the 0° angle of
attack, and under turbulent flow at the yaw angles of 0°, 5°, 10° and 15° with the 0° attack angle.
Taking as an example of Maanshan Bridge, the critical flutter speeds measured in these states are
listed and compared in Table 6. Under smooth flow, pure torsional flutter instability did exist in the
two testing cases, including State 2 at the attack angles of +3°and State 3 at the attack angles of −3°,
the corresponding critical flutter speeds are 74.2m/s (70.5m/s due to sectional model testing) and
79.9m/s (81.1m/s due to sectional model testing), respectively. For the rest states under smooth flow,
torsional and bending coupled flutter instability was observed with quite high critical flutter speeds,
from 90.2m/s to 100.4m/s, and the critical flutter speed under the 0° angle of attack from full
aeroelastic model testing is quite different from the result due to sectional model testing because of
different flutter modes. Flutter oscillation, however, did not happen in turbulent flow up to the wind
speed of 85m/s, beyond which the aeroelastic model stochastically vibrated with large amplitudes in
vertical bending but not related to any aerodynamic instability [13].
Table 6: Critical flutter speeds of Maanshan Bridge from wind tunnel testing
Yaw Angle of Critical Flutter Speed (m/s) Required
State Flow
Angle Attack Sectional Model Full Model (m/s)
1 Smooth 0° 0° 76.3 (Torsional) 90.2 (Coupled) 56.6
2 Smooth 0° +3° 70.5 (Torsional) 74.2 (Torsional) 56.6
3 Smooth 0° −3° 81.1 (Torsional) 79.9 (Torsional) 56.6
4 Smooth 5° 0° - 90.9 56.6
5 Smooth 10° 0° - 91.6 56.6
6 Smooth 15° 0° - 96.0 56.6
7 Smooth 20° 0° - 100.4 56.6
8 Turbulence 0° 0° - >85.8 56.6
9 Turbulence 5° 0° - >85.8 56.6
10 Turbulence 10° 0° - >85.8 56.6
11 Turbulence 15° 0° - >85.8 56.6

5. Conclusions
Bridging capacity and aerodynamic challenge of suspension bridges have been introduced with the
summary of ten longest suspension bridges, countermeasures of aerodynamic instability, super long
span suspension bridge and double main-span suspension bridges. With the experience gained from
ten longest suspension bridges, the intrinsic limit of span length due to aerodynamic stability is
about 1,500m for a traditional suspension bridge with either a streamlined box deck or a ventilative
truss girder. Beyond or even approaching this limit, designers should be prepared to improve
aerodynamic stability of the bridge by modifying cable system or adopting some countermeasures
on girder, including vertical and/or horizontal stabilizer and slotted deck as well as passive and
active control devices. Based on a preliminary study, either a widely slotted deck or a narrowly
slotted deck with vertical and horizontal stabilizers could provide a 5,000m span-length suspension
bridge with high enough critical wind speed, which can meet aerodynamic requirement in most
typhoon-prone areas in the world. The concept of double main-span suspension bridges for crossing
longer water body is an innovative way to increase bridging capacity with better structural static
and dynamic performance. Referred to dynamic characteristics, the lateral bending and torsional
vibration frequencies of double main-span suspension bridge are obviously greater than those of a
corresponding single main-span bridge although the vertical bending frequencies are almost same.
With this advantage, a double main-span suspension bridge has better aerodynamic characteristics,
in particular, high enough critical flutter speeds, which have been determined in the wind tunnel
testing with both sectional rigid model and full aeroelastic model.

Acknowledgement
The work described in this paper is partially supported by the Natural Science Foundation of China
under the Grants 90715039 , 51078276 and 51021140005, the Ministry of Science and Technology
of China under the Grants SLDRCE10-A-01 and SLDRCE10-B-05, and the Ministry of
Transportation of China under the Grants KLWRBMT-04.

References
[1] GE Y.J. and XIANG, H.F., “Aerodynamic Challenges in Long-Span Bridges”, Keynote Paper in
Proceedings of the Centenary Conference of the Institution of Structural Engineers, Hong Kong,
China, January 24-26, 2008. p.120-143.
[2] ASTIZ M.A., “Wind Related Behaviour of Alternative Suspension Systems”, Proceedings of 15th
Congress IABSE, Copenhagen, Denmark, 1996.
[3] MIYAZAKI M., et al., “Stay-Cable Systems of Long-Span Suspension Bridges for Coupled Flutter”,
Proceedings of 2nd European and African Conference on Wind Engineering, Genova, Italy, 1997.
[4] GIMSING N.J., Cable Supported Bridges, John Wiley & Sons, 1997.
[5] CHEN A.R., GUO Z.S., ZHOU Z.Y., MA R.J. and WANG D.L., “Study of Aerodynamic Performance
of Runyang Bridge”, Technical Report WT200218, State Key Laboratory for Disaster Reduction in
Civil Engineering, 2002. (in Chinese)
[6] GE Y.J. and XIANG H.F., “Chinese Major Bridges for Improving Traffic Infrastructure Nationwide”,
Keynote paper in Proceedings of IABSE Symposium 2007, Weimar, Germany, September 19-21,
2007, p.9-12.
[7] COBO DEL ARCO D. and BENGOECHEA A.C., “Some Proposals to Improve the Wind Stability
Performance of Long Span Bridges”, Proceedings of 2nd European and African Conference on Wind
Engineering, Genova, Italy, 1997.
[8] DIANA G., BRUNI A., COLLINA A. and ZASSO A., “Aerodynamic Challenges in Super Long Span
Bridge Design”, Bridge Aerodynamics, Larsen & Esdahl (eds.), Balkema, Rotterdam, 1998, p.131-
143.
[9] MIYATA T., YAMADA H. and DUNG N.N., “Proposed Measures for Flutter Control in Long Span
Bridges”, Proceedings of 15th Congress IABSE, Copenhagen, Denmark, 1996.
[10] ACHKIRE Y. and PREUMONT A., “Flutter Control of Cable-Stayed Bridges”, Proceedings of 2nd
European and African Conference on Wind Engineering, Genova, Italy, 1997.
[11] XIANG H.F. and GE Y.J., “On Aerodynamic Limit to Suspension Bridges”, Keynote Paper in
Proceedings the 11th International Conference on Wind Engineering, Texas, USA, June 2-5, 2003.
[12] GE Y.J. and XIANG H.F., “Bridging Capacity Innovations on Cable-Supported Bridges”, Keynote
Paper in the Proceedings of the 4th International Conference on Bridge Maintenance, Safety and
Management, Seoul, Korea, 2008, p.53-62.
[13] GE Y.J., XU L.S., ZHANG W.M. and ZHOU Z.Y., “Dynamic and Aerodynamic Characteristics of
New Suspension Bridges with Double Main Spans”, Proceedings of 7th Asian and Pacific
Conference on Wind Engineering, Taipei, Chinese Taiwan, November 8-12, 2009.