Gusset plates are large, steel sheets that hold together the steel beams of a bridge. Earlier this year the NTSB announced that one gusset plate, labeled U10, was underdesigned and was too thin.
Investigators began looking at the gusset plates surrounding that one. In it's recent finding, NTSB investigators said gusset plate L11 east was broken into five pieces. The plate on the west side broke into four pieces. Both plates were corroded and a crack followed the corrosion. Both plates were undersized, a problem that goes back to the original design of the bridge.
The investigators presented the evidence, but they did not answer a big question: Did the corroded, undersized gusset plate snap and lead to the bridge collapse or did another factor trigger the collapse and the plates ripped apart during the collapse?
"The NTSB will certainly address that," says Minnesota Department of Transportation Chief Bridge Engineer Dan Dorgan speaking on MPR's Midday program.
Starting in 1993 bridge inspection reports indicated rust and pitting in some gusset plates, he says. But these plates are large, and the corrosion on wasn't severe when the size of the plate is taken into account, he says.
"We believe that is what the inspector assumed in 1993 when they looked at it, so the corrosion since then may be additional, but the NTSB is certainly taking a thorough look at those issues; in fact some of the information they released addresses that. They'll continue to address that in their final report," Dorgan said.
Dorgan believes these findings indicate that the plate fractured after the bridge began to fall.
Analysis by NTSB investigators will determine that. The NTSB will likely use a scanning electron microscope to determine what kind of fractures occurred on this gusset plate, according to Donald Vannoy, a civil and environmental engineer at the University of Maryland.
"From a hypothetical point of view, if you looked at it you would see if this thing had cracked under some prior loads and was progressing. It's like it would crack and then it would get exposed to another overload and then it would crack a little more. And then one day the crack had progressed across so that it couldn't stand the load, versus something that is just ripped one time in a complete fracture," Vannoy says.
The NTSB is looking at gusset plate L11 because investigators believe the bridge collapse began nearby. Think of an upside down triangle. Gusset plate L11 is the peak. One arm connects to gusset plate U10 and the other U12. That triangle was one of several holding up the center span of the bridge. On August first that span also carried thousands of cars and nearly 300 tons of construction materials. "These are observational assessments without discussing cause," says Jerry Hajjar, a civil and environmental engineering professor at the University of Illinois at Urbana-Champaign.
It is very difficult to tell whether these cracks initiated before or after the collapse, he says. But he does have a tentative opinion.
"Having a little bit of corrosion on metal is probably not enough to precipitate something like this. That's just my feeling," he says. "It would have to be very significant corrosion or it would have to be coupled with something else that was significant such as a lot of extra load on the bridge than would be anticipated."
The NTSB will release its final analysis of the bridge collapse until mid-November. That report will detail what he calls, the "probable cause" of the collapse, says NTSB chairman Mark Rosenker.