The I-35W bridge across the Mississippi river at Minnesota collapsed without warning at rush hour lWednesday 1 August 2007, killing five and leaving a further eight unaccounted for.
Fatigue in the structures members is now thought to be the main focus of the investigation by the federal National Transport Safety Board (NTSB) which is now in full swing. It is currently carrying out a detailed finite element analysis of the bridge to help assess possible scenarios.
Minnesota’s state department of transport (MNDOT) this week confirmed that annual inspections of the bridge had relied on “visual inspections with load tests and strain gauges”. A spokesman said that there were no intrusive tests carried out.
But according to Mouchel Parkman’s structural technical director, Donald Pearson-Kirk who recently led inspection regimes in the US while with Parsons Brinkerhoff, visual inspections, such as those carried out on I-35W, would have missed crucial details.
“As well as looking, you have to do some testing, testing with some intrusive work, to get a much better picture," said Pearson-Kirk. "If something looks good, it might not always be so. Similarly, there are times when they look bad but are OK."
He said that with steel bridges, small incisions could be made and strain gauges installed to allow engineers to get a better picture, of precisely what was happening to the structure .Such techniques were not used at I-35W.
The bridge was a steel truss design built in 1967 with a concrete slab deck and was known to be structurally non-redundant. According to a recent report by consultants URS, written in 2006 it had 52 ‘fracture critical’ truss members along its 139.6m span (see page 6). The report was also critical of the inspection regime used to check the structure.
MNDOT this week confirmed that it would now review its inspection procedure and had engaged consultants Parsons Brinkerhoff to provide advise. It also said it would speed up inspections on five similar bridges. similar design.
Senior vice president of Chicago-based structural engineering firm CTL group, Gene Corley, also pointed out that a crucial piece of evidence must have been missed during inspections.
“Something was not being done,” he said, adding that if steel fatigue turned out to be the cause of the collapse "a visual inspection might not have picked this up."
However, MNDOT was keen to stress that it conducted inspections annually, when federal law required only biennial inspections.
The bridge was also undergoing non-structural resurfacing and joint replacement work at the time of the collapse (see below). Although the bridge is known to have been suffering from severe fatigue cracks and it is understood this work may have also been a contributory factor in the collapse
The resurfacing work closed two of the four lanes in each direction, but asymmetrically across the bridge span.
According to bridge expert Mark Whitby this "would give some interesting stresses in the two trusses – one going one way, the other the other way, and some strain in the cross-members.”
Another UK based structural engineer who wished to remain anonymous added: “The deck replacement work may have damaged a critical member of the truss. If there was a fatigue crack, who knows…they may have taken a piece of deck out above a fatigue crack, put a jack hammer on it and set it off."
He added: “It only needs one member in one truss to go for the whole thing to come down – truss bridges will not generally have any redundancy.”
Mark Raiss, chairman of Benaim, also said that the collapse is consistent with fatigue in the steel members.
“Fatigue always starts slowly, but you would expect cracks before anything began falling down," he said. “Most fatigue failures are preceded by visible cracks, but it is possible the crack length is small or invisible to the naked eye,” he said.
Ed Owen in Minnesota