Innovative Infrastructures - Toward Human Urbanism 1

Design Guidelines for Cable System of Long Span Bridges

Chan Min PARK Jong Gyun PAIK Sang Hoon SHIN

Managing Director Senior Manager Managing Director
COWI Korea Daelim Industrial Co., Ltd. CableBridge Co., Ltd.
Bungdang, Korea Seoul, Korea Seoul Korea
[email protected] [email protected] [email protected]

Ho Kyung KIM
Professor
Seoul National University
Seoul, Korea
[email protected]

Summary
Based on the active construction market of cable supported bridges, many experiences accumulated
in design and construction technologies as well as the development of high strength materials have
led the Korean engineer to think about new design scheme and concepts for long span cable
supported bridges. Under this background, Super long span bridge R&D center in Korea have
developed the reliability based design guideline for super long span bridges which allows 200 years
design life and is capable of using high strength cable and high performance steel as well as high
performance concrete.
Keywords: long span; cable supported bridge; high strength cable; high performance steel;
reliability based design; limit state; safety of cable; suspension bridge; cable stayed bridge; LRFD.

1. Introduction
Usually long span bridge is unique in its scale, boundary condition and public needs. And it also
requires longer life cycle while the maintenance work should be minimized due to the difficulties of
access and repair. For the past decades, many cable supported bridges are designed and constructed
in accordance mainly with the domestic bridge design specification which concept is limited to
medium size bridge. Consequently several foreign specifications and recommendations were
introduced case by case to fill up the lack of specific design manual for each long span bridge
project.
In order to improve this absurdity, ‘A Design Guideline on Cable Supported Steel Bridges’(KSCE,
2006), ‘A Guideline on using of Cable Materials in Long-Span Bridge’(KSCE, 2008) and ‘A
Design Guideline on Concrete Cable Stayed Bridge’(KICT, 2009) have been published successively.
Even though these efforts have contributed to resolve the thirsty for consistency and rationality in
design, these were a kind of agreement among domestic engineers based on some design examples
of advanced foreign countries and own experiences and knowledge without any research or detail
investigation.
In order to develop internationally competitive technologies, Super long span bridge R&D center
has been launched from May, 2009 as a government propelling program. Under the strategy of this
program, a reliability based design code (hereafter “Cable Bridge Code”) has been developed as one
of its key outcomes as well as high strength cable and high performance materials. And Cable
Bridge Code will cover these new developed materials. This code will allows the 200 year design
life for the primary structures with the corresponding reliability index of 4.0.
2 18TH CONGRESS OF IABSE, SEOUL, 2012

2. Design Concepts

Regarding the uniqueness of long span bridges, it is highly required to provide the reasonable
process how to prepare the loading and its combinations rather than just follow the deterministic
way by specifying certain coefficients, lack of variety. And, in order to keep the consistent
reliability level among the various elements, the concept of LRFD of AASHTO was used as the
basic philosophy. So, apart from the permanent loads and temperature effects, the other governing
loads were specifically studied and defined appropriately.

2.1 Limit state

The design requirements are based on classification of limit state design and implementation of the
various relevant load and resistance factors (or material partial factors) as specified in the new
Korean Bridge Design Code (hereafter ‘new KBDC’) or amended. All limit states shall be
considered of equal importance. And each component and connection shall satisfy Eq. 1 for each
limit state, unless otherwise specified.
∑ K iγ i Qi ≤ Rr (1)
in which:
Ki = importance factor relating to ductility, redundancy and operational importance
γi = load factor: a statistically based multiplier applied to force effects
Qi = force effect
Rr = factored resistance: φRn
( in case of concrete members, Rr =R{φi X i }, where X i is material strength )
φ = resistance factor: a statistically based multiplier applied to nominal resistance
Rn = nominal resistance

2.1.1 Serviceability limit state

The serviceability limit state is taken as restriction on stress, deformation, and crack width under
regular service condition. But if we think the large deformation characteristics of long span bridge,
dissimilarly to normal bridges, special consideration in load combination is in need. The followings
are the typical examples; (here, DL means dead load, LL live load, W wind, and T temperature)
For the movement of expansion joint: 1.0DL + 1.0LL + 0.6W + 0.T
For the navigational height: 1.0DL + .4LL + 0.5T

2.1.2 Strength limit state

Strength limit is taken to ensure that strength and stability, both local and global, are provided to
resist the specified statistically significant load combinations that a bridge is expected to experience
in its design life. The Cable Bridge Code has set the target reliability index, β =3.5 for normal cable
supported bridges and, β =4.0 for the important cable supported bridges. In order to fit this goal,
the importance factor, K i is introduced and multiplied as shown in equation (1) in case of strength
limit state. For all other limit states, K i =1.0
K i = 1.10 for important cable supported bridges
= 1.00 for typical cable supported bridge
≥ 0.95: relatively less important bridges
 Innovative Infrastructures - Toward Human Urbanism 3

This importance factor is introduced from the concept of Operational Importance in AASHTO
LRFD (1.3.5).
Similarly, in Euro Code, the consequence class CC3 and the reliability class RC3 EC0 recommends
a yearly maximum probability of bridge failure of 1 ⋅ 10 −7 . The corresponding accumulated
probability of failure of primary cable supported bridge elements or whole cable supported bridge
during 200 year lifetime of the bridge becomes 2 ⋅ 10 −5
All other structural elements than specified above shall meet consequence class CC2 and the
reliability class RC2 corresponding to a EC0 recommends a yearly maximum probability of bridge
failure of 1 ⋅ 10 −6 . The corresponding accumulated probability of failure during 200 year lifetime of
the bridge becomes 2 ⋅ 10 −4
Under these background, the target design life and its reliability index used in Cable Bridge Code is
shown in Table 1.
Table 1: Target Reliability
Structure Category Design life (year) Target reliability index

Long span Primary members 200 (or 150) 4.0

cable supported bridge 200 (or 150) 3.5
Secondary members
Small & medium span cable supported bridge 75 3.5
Replaceable elements (Bearing etc.) 25 3.5
Temporary bridge 10 3.0

2.1.3 Fatigue limit state

The fatigue limit state is taken as restrictions on stress range as a result of a single design truck at
the number of expected stress range cycles. K i and φ = 1.0 in the strength limit state equation
gives the fatigue limit state equation as follows:

γ (ΔF ) ≤ (ΔF )N (2)

γ = 0.75, load factor for the fatigue load combination which consider only fatigue truck load,
vehicular dynamic load allowance (IM=0.15) and vehicular centrifugal force
(ΔF ) = stress range due to the passage of the fatigue load multiplied by the factor 1.4
(ΔF )N = nominal fatigue resistance

The fatigue design truck load was developed for short and
medium spans. To account for the longer spans of cable
supported bridges the factor of 1.4 was developed in
accordance with the concept of PTI recommendation.
As the span length become longer, the effect of the weight of
the superstructure become bigger. In order to reduce the dead
weight of the steel orthotropic deck, the combination of high
Fig. 1: Fatigue Design Truck Load
strength steel and mild steel maybe required. Furthermore,
very thin layer pavement with the improved trough details
could be considered. In case of long span bridges, fatigue related specification could not be
deterministic in view point of cost efficiency and design life. It should be approached by experience
based theory and experiments to get the optimum details.
4 18TH CONGRESS OF IABSE, SEOUL, 2012

2.1.4 Extreme event limit state

This limit state is taken to ensure the structural survival of a bridge during a major earthquake, ship
collision and the rupture of stays in cable stayed bridges and hangers in suspension bridges.
Design of all cable-stayed bridges and suspension bridges should withstand the loss of any one
cable (stay or hanger) without the occurrence of structural instability.
For the rupture, the load combination and load factors would be 1.25DC + 1.50DW + 0.75(LL+IM)
+ Cable Loss Dynamic Forces. The impact dynamic force resulting from the sudden fracture of a
cable is of a magnitude twice that of the static force in the cable and acts at both the top and bottom
anchorage. The resistance factor φ =0.90 is suggested.

2.2 Main cable of suspension bridge

The stress of main cable may be considered to be composed of uniform axial stress from axial cable
force, bending flexure of the cable, transverse stresses at saddles and cable clamps, and uneven
axial stress due to built-in construction inaccuracies or temperature variation across main cable
section.
In usual design process, the main cable cross section is decided based on the maximum cable axial
force only. As a consequence, the resistance factor on the cable must be small considering the above
mentioned various additional stresses and possible deterioration of cable due to corrosion.
The main cable is designed for the load combinations in the ultimate limit state. The safety factor of
main cable is generally S=2.5. But the safety factors used in recent major suspension bridges are
less than it. Several examples are listed below with cable diameter( ), span length(L) and cable
strength.
- Akashi Bridge S=2.2 =1.1 m L=1,990 m 1,770 MPa
- Greatbelt Bridge S=2.2 =0.8 m L=1,620 m 1,570 MPa
- Messina Bridge S=2.1 =1.2 m (two) L=3,300 m 1,860 MPa
During the construction of Akashi Bridge in Japan, enormous studies and experiments were carried
out for the adoption of the high strength cable (1770MPa) and, as a consequence, S=2.2 was used.
In the near future, if we consider the improvements of material, water-proof technology, the safety
factor is supposed to be further optimized as it effects on bridge construction cost.
For the reliability based evaluation, various resistance factors are allowable in accordance with the
ratio of Dead load and Live load (Table 2).
Table 2: Reliability Based Resistance Factor for Suspension Bridge
Resistance factor, φ DL:LL ratio Safety Factor, S Reliability Index, β T
90:10 2.37 10.45
0.55 85:15 2.42 10.37
80:20 2.47 9.76
90:10 2.18 9.28
0.60 85:15 2.22 9.40
80:20 2.27 8.97
90:10 2.10 8.22
0.65 85:15 2.05 8.44
80:20 2.09 8.19
 Innovative Infrastructures - Toward Human Urbanism 5

In design of hanger, the generally used safety factor, S=2.5 and it is possible to find out the proper
resistance factor in accordance with the same approach used in main cable.

2.3 Stay cables of cable stayed bridge

Unlike suspension bridge, the stay cable is influenced largely by service load. So the serviceability
limit state would be checked.

2.3.1 Serviceability limit state

In general, the resistance factors for stay cable under the serviceability limit state are φ =0.45 in
service, φ =0.50 during construction. If the stay cable satisfy the fatigue test with the angle
deviation from the anchorage as specified in ‘A Guideline on using of Cable Materials in Long-
Span Bridge(KSCE, 2008)’, the resistance factors would be increased to φ =0.50, and φ =0.55,
respectively. These factors are evaluated on the assumption of using low relaxation material.

2.3.2 Strength limit state

Similar to suspension bridge, for the reliability based evaluation, various resistance factors are
allowable in accordance with the possible ratio of Dead load and Live load and corresponding
safety factor as shown in Table 2.
Table 3: Resistance factors for cable stayed bridge
Resistance factor, φ DL:LL ratio Safety Factor, S Reliability Index, β T
80:20 2.09 8.19
0.65 75:25 2.13 7.80
70:30 2.18 7.44
80:20 1.94 7.42
0.70 75:25 1.98 7.13
70:30 2.02 6.83
80:20 1.81 6.68
0.75 75:25 1.85 6.48
70:30 1.89 6.24

According to PTI recommendations (5th edition), for the limit states of AASHTO LRFD, the
following resistance factors shall apply:
- Strength A-Axial only φ =0.65
- Strength B-Axial and Bending combined φ =0.75

On the line of above investigations, safety factors of several suspension bridges and cable stayed
bridges in Korea were analysed and compared to some of international major bridges. In conclusion,
the allowed resistance factor for suspension bridge is in the range of 0.55~0.65, and 0.65~0.75 for
cable stayed bridge as shown in Table 2 and Table 3.
6 18TH CONGRESS OF IABSE, SEOUL, 2012

2.4 Cable Connection Details

2.4.1 Saddle
The main cable of suspension bridge passes through saddles at the pylon top and anchor blocks. The
key items to be checked in the design of cable at the saddle are to secure the friction between steel
wires and saddle groove and to keep the transverse load to any wire in the saddle not to exceed limit
in all load combinations of strength limit state.
Usually the coefficient of friction for a cable of galvanized steel wires passing through a zinc
sprayed saddle groove is taken as 0.2. And the corresponding resistance factor is φ =0.60.
Regarding the transverse pressure, it is well known that the value less than 500 kN/m does not
influence on the cable tensile strength and fatigue. And an experimental study carried out in Japan
shows that the effect of transverse pressure on tensile strength of cable with higher strength (1800
MPa) is less sensitive than the one with normal strength(1600 MPa). Thus if we assume the high
strength cable which has 1960 MPa or 2100 MPa it is necessary to adjust the limit of transverse
load. In fact, a couple of European bridges used the value of 600 kN/m.

2.4.2 Cable clamps

The friction between the main cable and clamps should be larger than the component of hanger
forces tangential to the main cable in all load combinations of strength limit state. This friction can
be acquired by clamping force which gives rise of radial pressure on the main cable.
The size of cable clamps (thickness, length, number of bolts and its force etc.) is determined
considering friction, stress of clamp and radial pressure on the main cable outer surface.
Usually the coefficient of friction for a cable of galvanized steel wires clamped by grooved zinc
sprayed casting is taken as 0.2. And the corresponding resistance factor is φ =0.60 considering the
loss of clamping forces due to relaxation of bolts and cable bend.
If the radial pressure is kept low, the clamp can be thinner but it should be longer. In this regards, as
an example, 3.6~6.0 MPa was used in USA and 16 MPa was used in the Severn Bridge. In general,
10 MPa has been used in Japan and recently European engineers tend to use 12 MPa (15 MPa for
central lock).
Through the experimental approach with high strength cables, above mentioned values are allowed
in actual design works. So it is necessary to select the design value considering the uniqueness of
each long span bridge.

3. Conclusions
The reliability based design guideline for super long span bridges which allows 200 years design
life has been provided as an outcome of vast investigation and research program. This program also
includes the development of high strength cable and high performance steel as well as high
performance concrete. Considering the uniqueness of the long span bridge, the guideline is
supposed to provide the process of determining each design value which allows rational and
consistent reliability as a whole structure. And the results were compared to the ones designed by
conventional deterministic approach.

4. References
[1] “A Design Guideline on Cable Supported Steel Bridges”, KSCE, 2006
[2] “A Guideline on using of Cable Materials in Long-Span Bridge”, KSCE, 2008
[3] “A Design Guideline on Concrete Cable Stayed Bridge”, KICT, 2009
 Innovative Infrastructures - Toward Human Urbanism 7

[4] LARS JENSEN et al., “Application of high strength steel in super long span modern
suspension bridge design, NSCC, 2009.
[5] PTI Recommendation for stay cable design - testing and installation (5th Ed.), 2007.
[6] FIB Recommendation for the acceptance of stay cable systems using prestressing steels,
2003.
[7] CIP Recommendations on cable stays, 2002