NTSB chief softens comments about bridge collapse

WASHINGTON -- The chairman of the National Transportation Safety Board backed off Monday from earlier remarks that seemed to rule out all but design flaws as the "critical factor" in the Interstate 35W bridge collapse.

Mark Rosenker's clarification, made to U.S. Rep. Jim Oberstar, D-Minn., came as NTSB investigators again contacted one of the researchers involved in a study of an Ohio bridge that buckled a decade ago due to undersized gusset plates similar to those that have become the focus of the 35W investigation.

Rosenker, meeting privately with Oberstar, said he did not mean to suggest that the finding of undersized gusset plates on the 35W bridge reflects the board's final conclusion on the cause of the Aug. 1 accident. That conclusion is expected later in the year.

Their meeting, and a separate letter to Oberstar, came a week after Oberstar criticized Rosenker for rushing to judgment on the cause of the collapse, calling his comments two weeks ago "highly inappropriate."

Rosenker, briefing reporters on Jan. 15, cited 16 undersized gusset plates that joined the 35W bridge's steel beams together as "the critical factor" in the collapse. He also said gusset plate design "tells us why the bridge collapsed."

In his letter to Oberstar, Rosenker noted that the NTSB still believes the bridge collapse began with the failure of one of the underdesigned gusset plates. He also reiterated that investigators still need to determine what caused the bridge to fail when it did.

Oberstar, chairman of the House Transportation and Infrastructure Committee, indicated through a spokesman that he is satisfied with Rosenker's response.

He and other Minnesota DFLers raised concerns about Rosenker's statements because they appeared to minimize the role of corrosion and maintenance in the bridge collapse. A design flaw would give them less of an opening to hold officials at the state Department of Transportation and Gov. Tim Pawlenty responsible for the bridge's upkeep.

There were also indications Monday that the NTSB is still reviewing a 1996 incident involving gusset plates on an interstate highway bridge near Cleveland for any links to the Minneapolis collapse.

Dean Palmer, a principal at Richland Engineering in Ohio, said a federal investigator contacted him Monday for more information on the incident, when gusset plates on both trusses on the eastbound I-90 bridge buckled, causing a 3-inch downward and 3-inch lateral displacement. Palmer was one of three experts who later analyzed what happened.

Palmer and another of the three experts, Arthur Huckelbridge, a civil engineering professor at Case Western Reserve University, said they were both initially contacted by the NTSB last fall about the Ohio bridge incident. Both experts said Monday that they did not know whether the Ohio incident was linked to the Minneapolis collapse, and said the NTSB had not indicated to them whether it saw a relationship.

"I talked to them [Monday]," said Palmer. "They were just looking for more information."

At his briefing two weeks ago, Rosenker said investigators were focusing on a possible design error involving gusset plates in Minneapolis but the collapse "is the only bridge failure of this type of which the Safety Board is aware."

However, Huckelbridge said an NTSB metallurgist contacted him Sept. 4 and wanted information on what occurred in Ohio. He said he assumed that the Minneapolis and Ohio incidents may not be similar because Rosenker had indicated that bridge inspections in Minneapolis, which had noted the corrosion of gusset plates, would not have identified a design error.

"[Corrosion] was definitely a factor on the Ohio bridge," he said. "So, if indeed corrosion was not an issue up on the I35 bridge, then there was some differences between the two."

In their report on the Ohio bridge, published in 1997, the authors said that a bridge painting contractor had a "line of trucks and heavy abrasive blasting, painting and lead-abatement equipment on the right-hand shoulder. A truck reportedly crossed the bridge in the left lane, causing the gusset plates to fail and sending painters scrambling to stable ground.

"More importantly, the gusset plates -- perhaps underdesigned to begin with -- had undergone corrosion that went completely through the plates in isolated spots. The corrosion had been hidden until prepaint abrasive blasting exposed it," the report stated.

mkaszuba@startribune.com • 612-673-4388 kdiaz@startribune.com • 202-408-2753