Bridge collapse: A half-inch closer to why

The collapse of the Interstate 35W bridge may have originated with the failure of gusset plates that were sized a half-inch too thin in the original 1960s design, the National Transportation Safety Board (NTSB) said Tuesday.

In a Washington news conference, NTSB Chairman Mark Rosenker also said there were no indications that Minnesota's upkeep of the bridge played any role in its collapse -- a statement that immediately unleashed political debate, with Gov. Tim Pawlenty scolding critics for leaping to conclusions and DFLers insisting that the full cause has yet to be found.

Rosenker emphasized that the federal probe is by no means final and that all the factors involved in the catastrophic collapse that killed 13 people and injured 145 on Aug. 1 are still months away from being known.

"We do not today know what caused the 35W bridge to collapse," he said.

The NTSB also warned bridge engineers nationwide Tuesday to review the designs and recalculate maximum loads on the estimated 12,600 similar steel deck truss bridges.

In Minnesota, Pawlenty ordered the Department of Transportation to recalculate the load capacities for 23 state bridges and help review them for 36 local bridges. All have designs similar to the I-35W bridge. Recalculations are underway for seven state trunk highway bridges, including the Hastings Bridge, with 16 to come.

Pawlenty said those recalculations will cost about $500,000 and should be complete by June.

For years, MnDOT bridge experts had worried about cracking and extensive corrosion on the fracture-critical I- 35W bridge, which they knew would collapse if any beam or connection broke.

Before the bridge collapsed, URS Inc., a private engineering firm, had been hired by MnDOT to study the span's structural integrity. The firm once noted that any buckling of gusset plates wouldn't be catastrophic.

Pawlenty said that the chairman's report "appears headed in a direction different than many of the political claims that have been made," and he called for his opponents to "reserve further judgment until the investigation is complete."

Democrats in Congress and DFLers in the Legislature argued that it is too soon in the investigation to rule out faulty maintenance as a factor.

Rep. Jim Oberstar, D-Minn., and chairman of the U.S. House Transportation and Infrastructure Committee, suggested that Rosenker is spinning the facts in favor of the Bush administration, which has opposed his plan for a national trust fund to repair decaying bridges.

"I think he got his tongue in front of his eyeteeth," Oberstar said. "To say that deterioration, rust and corrosion were not factors contradicts his other statements that no determination of probable cause has been found."

Thinned gusset plates

Gusset plates are metal plates that are installed where vertical, horizontal or diagonal beams are tied together with bolts or rivets. Two plates are used at each connection to sandwich the beams.

At the time it collapsed on Aug. 1, the I-35W bridge was holding some 300 tons of additional construction equipment involved in a repaving project. Investigators are looking closely at whether the extra load was a significant factor in the accident. Much of the equipment was concentrated over a node of gusset plates on the south side of the bridge's center span known as U-10.

Rosenker said it's possible that the U-10 connections, just below the concrete decking, were first to give way, but investigators are still months away from pinpointing what triggered the failure.

So far, investigators have found that of the 224 gusset plates, 16 at eight critical nodes in the center span were roughly half the 1-inch thickness required. All 16 of those plates were found to be fractured, while all other major gusset plates were generally intact.

The sizing error occurred in the original design process at Sverdrup & Parcel, a St. Louis firm that was later acquired by Jacobs Engineering. Because original work papers are missing, investigators don't know whether it was a math mistake or other miscue, Rosenker said.

He said no deficiencies have been found in the bridge's steel or concrete. Construction engineers would have built the bridge in strict accordance with the design specifications, he added.

Rosenker said gusset plates traditionally have been considered the strongest links in steel truss bridges. No one at the NTSB had ever heard of an underdesigned gusset plate, he said, and there's no reason to think similar design mistakes will be found in the upcoming sweep of safety reviews.

"It is important to understand that bridge inspections would not have identified the error in the design of the gusset plates," Rosenker said.

But MnDOT's history of inspection and study of the I-35W bridge before it collapsed was not without attention to gusset plates.

MnDOT's own bridge inspectors had been tracking corrosion in some of the gusset plates, particularly in one node known as L-11, one of the critical areas that is now known to have failed.

State inspectors found that a half-inch gusset plate there had lost nearly half its thickness in some spots because of corrosion, but no repairs were ordered, according to inspection reports.

Rosenker said he would not speculate whether reinforcing the undersized or damaged gusset plates would have prevented the bridge collapse. In the year before the I-35W bridge collapsed, URS studied the structural safety of the bridge. While the study focused on possible cracking in bridge beams, there was at least passing consideration given to the issue of gusset plate strength, documents show.

According to handwritten notes from a September 2005 consultants' meeting, a URS official wrote that "gusset plate buckling -- if this occurs, it is not catastrophic." URS has declined all requests for interviews on the I-35W bridge collapse.

Attorney James Schwebel, who represents two dozen victims of the collapse, took aim at URS on Tuesday, saying the firm was under contract to do computer modeling to determine the areas where stresses would be maximized.

URS "certainly had available all the same information the NTSB had available. Why didn't it spot this serious design defect and call it to the state's attention?" Schwebel said.

Improvements added weight

The I-35W bridge opened in 1967. In 1977 and 1998, MnDOT added weight to the deck by installing a center median barrier, building larger outside walls and increasing the thickness of the bridge's concrete deck by 2 inches.

"These changes added significantly to the weight on the structure," Rosenker said. "They were bringing the margins of error down to where they didn't exist anymore."

"This is the only bridge failure of this type of which the Safety Board is aware," Rosenker said.

MnDOT officials said they were bolstered by indications that the collapse might be attributed to a nearly 50-year-old design flaw and not to recent MnDOT actions.

"I think there is -- might be -- probably just a small measure of relief," said Dan Dorgan, the state bridge engineer.

Dorgan, flanked by Bob McFarlin, the assistant to Transportation Commissioner Carol Molnau, said that previous consulting studies on the I-35W bridge by the University of Minnesota and URS Inc. would have had no reason to focus on the gussets.

McFarlin and Dorgan said motorists had no reason to worry about traveling across bridges of similar design across Minnesota.

Minneapolis lawyer Chris Messerly, who belongs to a consortium of 20 law firms representing 67 victims of the bridge collapse, said Tuesday's events aggravated many of his clients.

"They [federal investigators] are pointing fingers at the designers, but the state knew or should have known something wasn't right with that bridge," Messerly said. "They dropped the ball for all the taxpayers."

He said he wants teams of private engineers to get access to the materials used in the investigation for an opinion separate from the conclusions that will be drawn by the NTSB and private investigations underwritten by the state.

Kevin Diaz reported from Washington, D.C. Paul McEnroe, Pam Louwagie and Patricia Lopez also contributed to this report. tonyk@startribune.com • 612-673-4213 kdiaz@startribune.com • 202-408-2753 mkaszuba@startribune.com • 612-673-4388