NTSB: Design flaw caused bridge collapse

January 15, 2008 7:32:13 PM PST
Steel plates connecting beams in the Interstate 35W bridge in Minneapolis were too thin by half and fractured, "the critical factor" in the collapse that killed 13 people and injured 145, the National Transportation Safety Board said Tuesday.

The connectors, called gusset plates, were roughly half the 1-inch thickness they should have been because of a design error, NTSB Chairman Mark Rosenker said. Investigators found 16 fractured gusset plates from the bridge's center span.

"It is the undersizing of the design which we believe is the critical factor here. It is the critical factor that began the process of this collapse. That's what failed," Rosenker said at a news conference.

What caused the bridge to collapse during rush-hour traffic in the early evening of Aug. 1 - "the straw that broke the camel's back," as Rosenker put it - was not yet known, he said. A final report by the NTSB was expected this fall.

The Minneapolis span was a steel-deck truss bridge that opened in 1967. Rosenker said it wasn't clear how the design flaw made it into the bridge because investigators couldn't find the design calculations.

The bridge was called "fracture critical," or lacking redundancies, meaning that a failure of any number of structural elements would cause the entire bridge to collapse.

Rosenker said the safety board had no evidence that the deficiencies in the Minneapolis bridge design "are widespread or go beyond this bridge."

However, the NTSB couldn't discount the possibility of similar errors in other like bridges, he said, and cautioned that states and contractors should look at the original design calculations for such bridges before they undertake "future operational changes." The NTSB issued a safety recommendation to the Transportation Department's Federal Highway Administration suggesting that the agency require bridge owners to do so.

Transportation Secretary Mary Peters called on states to calculate how changes in bridge weight, capacity or evolving bridge conditions will affect gusset plates.

"With a few calculations, we can help reassure travelers that our bridges remain safe," Peters said in a statement.

Rosenker noted that structural weight had been added to the Minneapolis bridge in two major renovations, in the 1970s and 1990s.

"When they added the weight they didn't realize they were bringing the margins of safety down to where they didn't exist anymore," he said.

Rosenker said that construction materials on the bridge the day it collapsed, which were part of a resurfacing project, added about 300 tons and were on the same side where failure of the bridge began.

Asked if the construction was the tipping point, Rosenker said, "I'm not ruling it in, and I'm not ruling it out." That will be left to the final report to determine, he said.

Abolhassan Astaneh-Asl, a professor of civil and environmental engineering at the University of California at Berkeley and an expert on gussets, spent 10 days in Minneapolis after the collapse gathering information for his own research.

"I concluded that the construction load may have been the last straw," he said.

Last August, Peters advised states to consider the additional stress placed on bridges during construction projects.

Rosenker said there was little chance that bridge inspectors would have noticed undersized gusset plates.

"No one recognized that you could undersize a gusset plate," Rosenker said.

"In the history of this organization, we have not seen anything like this before," he said, adding that gusset plates are supposed to be stronger than the beams they connect. "It was a shock, if you will, to the investigating team."

Dan Dorgan, the state bridge engineer in Minnesota, said that the assumption is that gussets wouldn't be the weakest point in the design.

"And, in fact, computer programs that generally model these bridges typically do not take gusset plates into consideration," Dorgan said. "They mainly look at the main members."

The Minneapolis bridge was deemed "structurally deficient" by the federal government as far back as 1990, and the state's maintenance of the structure has been questioned. But Rosenker said the NTSB investigation has found no evidence that cracking, corrosion or other wear "played any role in the collapse of the bridge."

Investigators also found no flaws in the steel and concrete material used in the bridge.

The bridge was originally designed by Sverdrup & Parcel, a company acquired in 1999 by Jacobs Engineering Group Inc. of Pasadena, Calif. A message left by The Associated Press with Jacobs Engineering wasn't immediately returned.

Minnesota Gov. Tim Pawlenty noted that he had warned others not to jump to conclusions about the collapse. Pawlenty and his transportation commissioner, Lt. Gov. Carol Molnau, have been sharply criticized over state funding for infrastructure and maintenance.

"While the NTSB investigation is not complete, the focus of the investigation appears headed in a direction different than many of the political claims that have been made," Pawlenty said.

Minnesota is reviewing 23 state bridges with truss designs to make sure their current load ratings fit with the gusset design. Officials hope to complete the state reviews by June at a cost of $500,000. There are another 36 such bridges controlled by municipal governments in Minnesota.

Late last year, President Bush signed a massive spending bill that included $195 million to help replace the bridge. That came on top of the $178.5 million the federal government had already given Minnesota for the project.

Associated Press reporters Brian Bakst in St. Paul and Steve Karnowski in Minneapolis contributed to this story.