As investigators sort out the cause of the Aug. 1 I-35W bridge collapse in Minneapolis, one clue to the disaster may be in the design itself.
As crews clean up debris from the collapse, National Transportation Safety Board (NTSB) investigators said they found a potential design problem with the gusset plates, steel plates that tie together angled steel beams of the bridge’s frame.
Those plates are used to help hold the steel trusses together.
Investigators are trying to verify what loads and stresses these plates faced at specific locations, as well as the materials used to construct them.
One possible stress was the weight of construction equipment and materials on the bridge when it collapsed, the U.S. Transportation Department said in a statement Aug. 8. Trucks, loaders and paving equipment rumbled onto the Interstate 35W bridge in mid-June as Progressive Contractors Inc. (PCI) began a $2.4 million project to repair sections of the heavily traveled highway.
Safety board and transportation officials said the increased focus on steel plates is preliminary. They would not say exactly where they were located on the bridge or whether their failure alone would have caused the collapse.
The list of possible causes remains lengthy in the early stages of the investigation. It includes a lack of redundancy in the bridge’s design, aging steel, rotting welds, vibrations from adjacent train tracks and even the corrosive effects of bird droppings.
“We are continuing to make progress on this investigation, and each area of inquiry gets us closer to ultimately determining the cause of this tragedy,” Safety Board Chairman Mark Rosenker said in a statement.
Federal transportation officials were concerned enough with the NTSB finding to issue an advisory to states to “carefully consider” the weight of construction equipment used in bridge projects.
Charles Walczak, president of the Structural Engineers Association of South Carolina, said the gusset plates, which can be worn through fatigue, could have been cracked or may been made from defective steel.
“There may have been some slag used from the ingot of steel,” he said. “In the [manufacturing] process, there may have been some defects in rolling out the materials.”
If it was fatigue, it may have been a matter of time and wear.
“The properties of steel change through stresses,” he said.
Fatigue, however, is something inspectors look for when conducting reviews of these spans.
“That is what inspections are for,” Walczak said.
And the deficient rating that bridge received during the recent inspections, he said, should set off some warning signs.
The bridge, which opened in 1967, was a steel truss bridge consisting of three parts: the deck, superstructure and substructure. It had been inspected annually since 1993 and every other year prior to that. It was scheduled for reconstruction in 2020 to 2025, according to the Minnesota Department of Transportation.
After the bridge collapse, which attracted world attention and brought visits from President George Bush and First Lady Laura Bush, $250 million was earmarked to replace the span.
A Search for the Cause
The 1,907-ft. bridge collapsed into the Mississippi River toward the end of a PCI shift, taking the 18-person crew with it. At least 9 people were confirmed dead and approximately 100 were injured. It was unclear at press time if searchers had found one PCI crew member reported missing.
Employees of PCI, based in St. Michael, Minn., have helped investigators map out the locations of its equipment, vehicles and materials at the time of the accident, and how much each piece weighed. The company was working on a section of bridge over the river toward the southern end, said NTSB chairman Mark Rosenker.
PCI lawyer David Lillehaug said there is no reason to think the project had anything to do with the collapse. The most recent jackhammering of deteriorating concrete had occurred the day before, and none was performed the day of the accident, he said.
Another question for the investigators is whether the diversion of traffic caused instability among the supporting structure’s steel arch ribs. The project took up four lanes of the eight-lane bridge.
The contract required PCI to repair the bridge deck, replace the concrete surface and expansion joints and work on the anti-icing system, said Minnesota Department of Transportation spokesman Kevin Gutknecht.
Company officials said the crew was preparing to pour a 2-in. (5 cm) layer of concrete when the span gave way.
Both Gutknecht and Lillehaug declined to provide additional details about the project.
The work was being done in segments, using industrial saws to cut around damaged sections of concrete, then using jackhammers to remove the severed chunks, said Glen D. Johnson, business manager of International Union of Operating Engineers Local 49, whose members were among the PCI crew.
“There were hunks in there they just basically chip out. They jackhammer it out, smooth it off and put a new coat on,’’ said Johnson, who had visited and driven past the job site regularly, most recently the day before the collapse.
PCI, founded in 1971, does approximately $100 million in annual business. It works primarily for state and local governments in Minnesota and other states in the Upper Midwest, specializing in road paving and repair, bridge reconstruction and steel fabrication.
A Range of Theories
Bridge experts have a range of theories about whether the actual construction on the bridge could have played a role in the collapse.
One who discounts a connection is structural engineer Stuart Sokoloff, owner of CTS Group in New York, a company that performs engineering failure analysis.
“If indeed the work that was going on was only patching or resurfacing, even if they’re using jackhammers, I can’t see how that would cause a major catastrophic failure of these massive [steel] trusses,’’ said Sokoloff. “I just don’t put the two of them together.’’
Gianluca Cusatis, assistant professor of civil engineering of Rensselaer Polytechnic Institute, in Troy, N.Y., who specializes in structural design, said the steel arch design of that bridge is problematic because it may lack support redundancies in case a weight-bearing member suddenly collapses.
“There’s no stability,” he said. “There’s no possibility to redistribute the load. [The design] is safe as long as nothing goes wrong.”
Although the bridge was undergoing some deck improvements at the time of the rush hour collapse, Cusatis thinks it’s unlikely that work would have had anything to do with the accident.
“It had nothing to do with it,” he said. “I really don’t think the construction on the deck would affect it any way. The deck did not cause the collapse.”
Instead there’s a possibility that despite passing a recent inspection, the 40-year-old bridge may have suffered structural fatigue or corrosion problems that, when combined with a lack of secondary supports, could have triggered the calamity.
A report released in July 2006, after an inspection a month earlier, found the bridge’s superstructure was susceptible to fatigue in the truss and floor members. These problems raised questions about consequences surrounding the failure of a main truss member that could be triggered by a fatigue crack.
The bridge was rated a four, or deficient, on a scale of zero to nine, with a zero-rating as the worst possible condition.
The bridge, which opened in 1967, was designed with standards set in 1961, but the report relied on new standards that were revised in 1974. That report is the basis for evaluations by the Minnesota’s Department of Transportation to determine the future renovations of the bridge.
Also, the data were being used to help identify critical superstructure members prone to cracking and the evaluation of structural consequences during the failure of a critical member. The report also was a basis for the state agency to develop contingency repairs for the fracture of critical members, to improve the structural redundancy, reduce truss stress and develop a deck replacement plan.
The 2006 report found that five main truss members were prone to fractures and should be retrofitted with high-performance steel and high-strength bolts. The retrofit, according to the study, would enhance the redundancy to truss members. And the deck replacement on the main truss spans would reduce live load stresses and enhance the redundancy of the truss system.
Corrosion alone, said Cusatis, could be a culprit if it compromised a main load bearing support. If that support collapses and there are no built-in redundancies, the bridge could come down.
“It slowly deteriorates the carrying capacity of a support member,” he said.
And that support member, especially under rush hour stress, is unable to carry its load at its most needed time. To prevent disasters of this kind, Cusatis said these bridges are designed now with secondary support systems.
“When we design,” he said, “we make sure there is some kind of redundancy. We need to make sure members don’t collapse.”
But the lack of these built-in supports doesn’t necessarily mean these steel-framed bridges are hazardous, added Cusatis.
Some of the steel truss bridges designed and built during the 1950s and 1960s, such as the I-35W bridge, are considered somewhat obsolete due to their lack of structural redundancies.
Walczak of the Structural Engineers Association of Carolina said if a support did give way, it could have caused the disaster.
“It could have been a chain reaction collapse,” said Walczak, a native of upstate New York who served as the engineer-in-chief during the construction of the south span of the Newburgh-Beacon Bridge, which crosses the Hudson River on I-84. “It’s possible that’s what happened.”
The Advent of Fed Inspections
Walczak graduated from college in the late 1960s, and he entered his chosen field at the time to work for the New York State Department of Transportation when another bridge disaster stunned the nation.
In the middle of the Christmas shopping season, a 1,750-ft. (533.4 cm) span known as the Silver Bridge, due to its color, which crossed the Ohio River from Point Pleasant, W.Va., to Kanuaga, Ohio, suddenly collapsed at 4:55 p.m., Dec. 15, 1967, killing 46 people and injuring hundreds of others.
The two-lane suspension span was built in 1928 and was packed with traffic that day. Witnesses said it started twisting on the north side, on the Ohio border, before turning over like a deck of cards by the time it reached West Virginia.
An estimated 60 to 70 cars were packed onto the bridge across U.S. Route 35 when it started bending at the north side. It spilled vehicles into the cold waters of the Ohio River and then fell onto them, witnesses said.
The suspension bridge was similar to the Golden Gate Bridge in San Francisco and the Verrazano-Narrows Bridge in New York City. But instead of spun-wire cables used to hold up the support towers on those bridges, the Silver Bridge was held up by carbon steel chains that were anchored on each shore of the river.
A National Transportation Safety Board report found that a joint supporting a chain snapped on the north side, toppling both towers and sending the bridge into the water.
The bridge had no load limit and 37 vehicles were on the Ohio side, including five tractor-trailers and a pair of gravel trucks. The heavy load brought the north side down and facilitated the immediate demise of the south-side after the chains ripped loose.
A thorough inspection of the bridge had not been conducted for 16 years prior to the disaster, and inspectors sometimes used binoculars to inspect the chains from the road deck. The bridge’s structure also suffered from corrosion problems.
Within a week of the Silver Bridge disaster, Sen. Jennings Randolph, D-W.Va., who chaired the Senate’s Public Works Committee, called for hearings, which eventually led to requirements for the first federal bridge inspection requirements.
“A lot of bridges were rehabilitated after that,” said Walczak. “I’m sure something is going to happen again now with this tragedy.”
(The Associated Press contributed to this article.)CEG