Bridge over troubled water
Sometimes a glass of water can cause a lump in your throat. In an attempt to get a feel for the ocean floor of the Northumberland Strait in Atlantic Canada, a small piece of mudstone was delivered to J. Muller International in San Diego. After passing through a few hands, Vice President of Engineering Gerard Sauvageot plopped the specimen into a glass of water.
“It was pretty hard, you could scratch it with a knife, and it seemed like pretty good rock,” Walter Eggers, senior bridge engineer at J. Muller International, told ROADS & BRIDGES. “Gerard took that thing and he got a glass of water and put it in the glass of water, and within 15 minutes it was dissolved to nothing. That had a pretty strong effect on all of us. We were very concerned about being sure that in the end we would have a confident material underneath the foundation.”
Sauvageot’s simple experiment only magnified the fears of a project tagged virtually impossible due to the incredible forces of nature standing between the Provinces of Prince Edward Island and New Brunswick.
But conducting tests also led to several innovations which now comprise the 13-km-long Confederation Bridge, this year’s winner of the 1999 George S. Richardson Medal which was presented to Sauvageot at the International Bridge Conference in Pittsburgh on June 13. The award recognizes a single, recent outstanding achievement in bridge construction. The Strait Crossing Joint Venture, comprised of SC Infrastructure (formerly SCI of Canada), Morrison Knudsen of the U.S., GTM of France and Ballast Nedam of the Netherlands, was the contractor for the bridge. J. Muller International and Stanley Consultants designed the project.
“This is a feeling of accomplishment throughout the whole company,” Daniel Tassin, senior executive vice president for J. Muller International, told ROADS & BRIDGES. “Nobody totally believed that it could be done within the schedule that was set, and eventually it was done and we had very, very, very few problems at the site for such a huge structure and such an unusual structure.”
In the beginning, however, the problems facing everyone involved with building one of the longest concrete bridges in the world were fierce.
‘Freezing to death’
For years a ferry was the only mode of transportation between Prince Edward Island and New Brunswick, and Eggers won’t soon forget the time he saw one trying to break through the ice in the dead of winter. It was about 20 below zero, the wind was gusting at approximately 30 mph, little flecks of ice were blowing up on shore and Eggers was trying to make sense out of placing a concrete bridge through a thick sheet of ice.
“There was packed ice all the way across the ocean, and I was just freezing to death,” recalled Eggers. “I couldn’t believe anybody was going to build a bridge across that strait. It was a horrible place to be . . . I thought they were all crazy.”
And after seeing the design requirements issued by Public Works Canada, Eggers must have been questioning somebody’s state of mind. Tough standards, however, made the bridge what it is today: a tremendous feat.
The most important of the requirements were as follows:
• The facility should provide a service life of 100-plus years;
• Consideration must be given to fundamentals of aesthetics;
• Environmental loads such as ice, wind, wave and current, earthquake and temperature must be taken into account;
• The structure would not be prone to progressive collapse;
• The structure would be able to withstand a certain magnitude of ship collision;
• A 172-m-wide navigation channel with 49-m vertical clearance and 13-m water depth would be provided; and
• The roadway width would be able to accomodate three traffic lanes.
“What we really ended up having to do was this side of the impossible, at least in everybody’s mind,” Ross Gilmour, director of design and engineering for SC Infrastructure, told ROADS & BRIDGES. “It was like enlisting in the army, you sign yourself up for a tour. You live it, you die it, you go to bed with it, you wake up with it and go back to try to build it the next day.”
The bridge vs. the elements
Ice, wind, progressive collapse, ship collision and the ocean floor were all key figures in determining what kind of bridge would be erected, and the contractor and bridge designers executed several tests and formulas to come up with solutions.
Ice is present in the straight for about five months out of the year, with the worst conditions occurring in late March. Maximum ice flow thickness is 0.30 m with a typical floe diam of 120 m.
Wave, wind and current break the floes and form ice ridges, each one generally made up of a core of consolidated blocks and refrozen ice at the water line, with a small sail of loose ice blocks above and a much larger keel of partially consolidated ice rubble below. Ice ridges may reach 50-75 m in diam with a core thickness of 2.5 m.
To help crack the ice, the bridge designers developed a 52 degree conical ice shield located on the pier shaft. The idea behind the ice shield was to present an inclined surface to the oncoming sheet of ice, so instead of having a crushing failure of the ice, which generates a very high load, the shield actually lifts the ice and makes it break under its own weight through bending. Concrete for the ice shields has 90 MPa (megapascal) at 28 days for resisting ice abrasion.
“The ice shield isn’t a new idea. We’ve used it before ourselves back in the ’70s,” said Eggers. “They were having big concerns about the ice loads before we even signed the contract. The ice shield also had the added benefit that it was so big and heavy that it just added a lot of dead weight to the gravity-based foundation. So it did help provide stability.”
Reference wind speed is 26.5 m/sec for a 100-year return period, with a reference wind pressure of 0.44 kPa. Two basic types of tests were done on the design in a wind tunnel: sectional and aeroelastic. The sectional tests determined the aerodynamic characteristics of the cross section, while aeroelastic tests were conducted to determine the response of the bridge to turbulent wind. The aerolastic tests, reserved for cable-supported bridges, showed that the dynamic response of the bridge-to-wind was much greater than anyone had expected.
Because of the test results, the reinforcement in the piers had to be increased and some of the foundation elevations were adjusted.
In the event of a collapse of one pier or loss of one span, the structure was built in a way that such a collapse would not trigger the collapse of the entire bridge. Rigid frames with alternating drop-in spans were used to deal with the potential threat, and in the event of a ship collision the piers next to the navigation channel were designed to withstand the impact of a 37,000 DWT (dead-weight tonnage) vessel at a speed of 4.2 m/sec. The two adjacent piers can withstand an impact force of 50 MN (meganewtons) at mean high water level. All other piers were designed to survive an impact force of 8 MN.
The bedrock sequence across the strait consisted of a series of interbedded sandstone, siltstone and mudstone layers overlain by up to 14 m of glacial till. The foundation level for the pier bases was chosen on competent sandstone, with the lowest foundation 39 m under water. To help deal with the mudstone, tremie concrete was poured underneath the pier base to match up with the questionable surface.
“That gave you a real solid contact and it also capped off the mudstone and sealed it so water wouldn’t get in there,” said Eggers.
The foundation level for each pier was dictated by the competent sandstone level with regard to vertical load capacity and horizontal sliding resistance. Several standard pier bases of different heights were designed and used according to the foundation levels.
Spread footings shaped like doughnuts were used at the base. The circler design turned out to be very efficient in dealing with both tranverse and lateral longitudinal loads in the direction of the bridge, according to Gilmour.
“You had very large forces in both directions, so the circle ended up being efficient,” he said.
To dredge the spread footings, a special dredging template was used to perform certain functions like guiding the dredge buckets.
Diving right in
Only four components—pier base, pier shaft, cantilevers and drop-in span—weighing 7,800 tonnes were used to assemble the 43 main bridge spans of 250 m each. Precasting of the pier bases, pier shafts, cantilevers and drop-in spans was performed on Prince Edward Island, while the approaches, segments for the piers and the superstructure were prefabricated on a yard on the New Brunswick side.
The maximum diameter of the pier base was 28 m, the maximum height of the pier shaft was 48 m and the maximum lengths of the double-cantilever deck and drop-in span were 192 and 64 m, respectively. Concrete for the bridge has properties of 55 MPa at 28 days and 60 MPa at 90 days minimum.
The precasting began in 1994, with the actual placement of the pier bases coming in August of 1995. Once all the precasting was complete, the piecing together of the bridge took just 12 months.
The four main components were transported from the Prince Edward yard on a Huismann hydraulically operated sledge and carried to the site by a catamaran-type of floating crane, called HLV Svanen.
The pier base was set on hardpoints in a trench dredged into sandstone, then tremie concrete was placed between the base and the sandstone.
The pier shaft was grouted and post-tensioned onto the pier base, and the main cantilever, with the underside of the pier segment matchcast to a template placed on top of the pier, was set in place and post-tensioned to the pier.
A drop-in span was then inserted between the tips of two adjacent cantilevers, acting as a strut between the structures during its placement, and connected by a means of narrow closure pours and post-tensioning tendons which transformed the whole into a continuous moment-resistant frame.
A hinged drop-in span was placed between adjacent frames.
For the approaches, piers were precast segmentally and assembled with post-tensioning tendons. The 93-m spans were made up of 15 pairs of precast segments placed by the balanced cantilever method. Segments were set in place by a twin truss launching girder, leased by an Italian company called Beal, resting on the pier where the balanced cantilevers were erected.
“This was one of the most complex jobs that we’ve ever worked on,” said Tassin. “The structure was complex, the substructure was complex and coming up with an innovative design in a very short time frame was difficult.”
The right idea
The intense nature of the project required some intense thinking, and with that came some impressive ingenuity.
Since the Ontario Highway Bridge Design Code was developed for a 50-year design life, a new set of codes had to be set for one expected to last 100-plus years. Dr. James McGregor, a past president of the American Concrete Institute, was hired to develop a new code for the load and resistance factors.
A calibration process using probabilistic reliability techniques was used, and the design target safety index ß, which is a measure for the probability of failure of a structure member, was 4.0 for multi-load path and 4.25 for single-load path components.
Tassin came up with a revolution while in the shower.
Trying to figure out how to erect the main girder, Tassin got the idea of building a precast, matchcast template that would allow them to erect a tiny piece of concrete, weighing about 100 tons, up on top of the pier shaft so the piece could be surveyed and set accurately. This eliminated any surveying done to the main girder when it was put in place.
Sauvageot devised the plan to install the drop-in girder between the main girders without counterweights, thus saving time and money. Instead of putting a counterweight on the other side of the cantilever to balance the weight of the drop-in, Sauvageot eliminated the counterweight by using the drop-in span as a strut.
In addition, the tranverse post-tensioning for the deck slab was all done with high-density polyethylene ducts instead of corrugated metal ducts.
“In trying to convince ourselves that it could last a hundred years, we probably built it to last 200,” said Gilmour.” We remain optimistic, you have to be to be in this business.”