Suspension Bridge Records · South Korea

Sorok Bridge: How It Was Built

A self-anchored suspension bridge of 110 + 250 + 110 m with a single central cable, linking the Goheung mainland with Sorokdo, opened to traffic in 2009.

Sorok Bridge was the first stage of the road from Doyang on the Goheung mainland via Sorokdo to Geogeumdo; Geogeum Bridge later carried the route on to Geogeumdo. Its 470 m suspension section is Korea's first mono-cable self-anchored suspension bridge and was described as only the second of its kind in the world. A single main cable runs down the centre line between diamond-shaped concrete towers and is fixed directly to the steel deck, with no anchor blocks. With its connection bridge the crossing is 1,160 m long.

The bridge is small by suspension bridge standards, with a main cable just 300 mm in diameter, which makes it a clear example of the two things that set self-anchored bridges apart: how the deck and cable are erected, and what the cable needs over its life.

Key facts

Design

The design and construction of Sorok drew on Korea's earlier long-span bridges, including Yeongjong and Gwangan. With no anchor blocks, the steel box girder carries the horizontal pull of the cable, and the single cable in the centre line keeps the deck free of cable planes at its edges. The bridge was designed for a wind speed of 56.1 m/s and used 3,900 t of structural steel, 283 t of cable steel and 9,413 m³ of concrete.

The main cable was made from prefabricated parallel wire strands rather than spun on site. The engineers chose PWS over air spinning because it is faster, quality is easier to control, the equipment is simpler and it needs less labour. Fourteen strands of 127 wires make up the 300 mm cable.

For the erection stages the designers compared two ways of transferring the deck from its temporary supports onto the cable: jacking the girder down, or pulling in the hangers. They chose hanger pulling for stability and cost. Main cable and hanger forces were solved together to keep the girder camber small, and the girder profile measured on site matched the erection-stage analysis.

How it was built

Pulling out the strands

The 14 strands were erected in the autumn of 2006. Each strand came on a reel of about 21 t, delivered by trailer in erection order and lifted onto a 40 t unreeler at the P7 end by a 50 t hydraulic crane. The strand socket was hauled along a tramway: a carrier ran on a support rope above the catwalk, pulled by two-drum winches of 5 t at the tower tops and at the anchorage, with 2.7 t cranes on the tower tops. At each tower the socket was lifted through the tower-top bent with a 5 t chain block, the carrier was transferred, and the strand was pulled across the main span by the two tower winches, one hauling while the other paid out.

Keeping the strands straight

Each strand has a coloured marker wire that shows any twist. The marker was set at the 11 o'clock position and checked in every span; twists picked up at the tower-top rollers were turned back with a chain wrench and a wire aligner before the strand was set. The record lists the problems met and their cures: seizing tape bursting as the strand left the reel (solved by unreeling at a steady speed and re-taping at once), wires slipping off rollers in the side spans (solved by posting a watcher at each roller), and the inner turns of each reel arriving most disordered, with the marker wire rotated by 90 to 180 degrees.

Into the saddles

Once a strand hung free on temporary clamps, it was reshaped from its hexagonal section into a square with forming clamps so that it would fit the saddle, laid into place, and held apart from the next strand by wooden wedges. Wooden and plastic hammers were used to free strands that stuck in the saddles.

Sag adjustment at night

Strand No. 1 was the reference. Its absolute sag was surveyed at night, when the strand temperature was uniform, with a Geodimeter and prisms at the saddles and midspans; the absolute survey was repeated after strands 7 and 14. The strands in between were set in four rounds of relative sag measurement: the height difference to a reference strand was measured with calipers and a spirit level, four contact thermometers gave the strand temperatures, and the reading was corrected for the temperature difference. The tolerance on relative sag was plus or minus 30 mm. The sag was sensitive: about 3.6 mm per degree in the main span and 4.6 to 4.7 mm per degree in the side spans, while 1 mm of strand length moved the main-span sag by about 1.1 mm and the side-span sag by about 2.9 mm. Adjustments were made at one tower, PY2, with a strand jack and clamp, while the other tower top, PY1, was the fixed point.

At the anchorages each socket was first fixed so that the strand hung 100 to 200 mm above its target sag, leaving room for the night adjustment. In the anchor spans, where sag is hard to adjust, the strands were set by tension instead: a centre-hole jack pulled the socket while socket movement and jack pressure were logged, and a regression line through at least five points gave the force in the strand, which was brought to within plus or minus 0.2 t of its target.

Load test and opening

Before opening, a 30 t truck was parked at the side-span midpoint and at the quarter, three-eighths and midpoints of the main span, then driven across at 5 to 70 km/h while differential GPS recorded the deck at 10 Hz. The bridge opened to pedestrians for five days in September 2007 and to traffic on 2 March 2009, ending the ferry service to Sorokdo.

Key dates

What the cable bands taught

The 33 cable bands are clamped to the cable with 14 to 22 bolts each, and friction between band and cable keeps them from sliding. In September 2013, about four and a half years after completion, every band was checked. The bolts had lost on average 35.4 percent of their initial force, and 24 of the 33 bands had fallen below the slip safety factor of 3.5 required by the Korean specification, taking the band-to-cable friction coefficient as 0.15. Ordinary bolt relaxation accounts for only 2 to 5 percent; the authors traced most of the loss to creep of the zinc coating on the cable wires under the band pressure.

All bands were re-tightened over about two months, raising the average bolt force from 49.0 tf to 72.7 tf and the safety factor to 4.0 or more. Monitoring afterwards showed the bolts kept 86.4 percent of their force after 331 days, a slower loss than after completion. Yeongjong Bridge had shown the same pattern, and the case is a practical reminder to plan band bolt checks and re-tightening into the maintenance of any suspension bridge.

Construction records

The records include the Korean PWS cable erection record (as PDF and as the original HWP document) and the Korean construction record of the bridge.

← All suspension bridge records

Related on TheBridgeEng

Questions engineers ask

What kind of bridge is Sorok Bridge?

A self-anchored suspension bridge with a single central main cable fixed to the steel deck, with spans of 110 + 250 + 110 m. It was Korea's first mono-cable self-anchored suspension bridge.

How was the main cable of Sorok Bridge erected?

From 14 prefabricated parallel wire strands of 127 wires, pulled across by a tramway and winches in the autumn of 2006. Strand No. 1 was set by absolute sag survey at night, and the others by relative sag against a reference strand.

Why did the Sorok Bridge cable band bolts lose force?

Mainly through creep of the zinc coating on the cable wires under the band pressure. After about 4.5 years the bolts had lost 35 percent of their force on average, and all 33 bands were re-tightened in 2013.

Sources

• Structurae: Sorok Bridge
• Structurae: Sorok Bridge literature list
• Bridgemeister: Sorok (2008)
• CERIK, Lee Deok-su column 112: Sorok Bridge, Korea's first mono-cable self-anchored suspension bridge
• Son et al., Analysis for Initial Equilibrium Condition and Erection Stages of Sorok (Self-Anchored Suspension) Bridge, 17th IABSE Congress, Chicago 2008 (Structurae record)
• Kim et al., The Recent Cable and Deck Erection Methods Applied in Various Types of Suspension Bridges in Korea, IABSE Geneva 2015 (Structurae record)
• Chun, Choi and Kim, Sorok Bridge: Standard for an Economical Suspension Bridge, SEI 22(1) 2012 (Structurae record, no abstract)
• Seo, Kim and Kim, DGPS displacement measurement of a self-anchored suspension bridge, KSNVE Spring Conference 2009
• Korean Wikipedia: 소록대교
• Korean Wikipedia: 국도 제27호선

Updated 2026-11-13 · written by a practising bridge engineer