Infrastructure upgrades are expensive and inconvenient - you've probably heard of the Big Dig in Boston. They take a long time to complete - the most recent water tunnel for NYC (Tunnel no. 3), they've been digging for 40 years and won't finish until 2020. Definitely not sexy. Not like building a new baseball stadium.
Here in Seattle, we've been at loggerheads about replacing our aging, sinking, earthquake-addled double-decker concrete highway (the Viaduct) ever since the 1989 quake collapsed a similar design in SF/Oakland. After the Nisqually Quake in 2001 toppled buildings below the Viaduct the debate became reinvigorated. Of course debate does not equal action. So we continue to close it for inspection twice a year, and bolt more steel to the columns to keep it standing.
So anyway, we have old highways, old levees, old steam tunnels, old sea walls, old reservoirs, old dams, old power transmission systems, and yes - we have old bridges. It would have been nice to have spent over the last 7 years billions of dollars on aging infrastructure in this country. I can think of a few cities that really could have used the help. But I suppose that will have to wait for a different administration with different ideas about what it means to "secure the future for America's children".
In the meantime, here's some real news about the bridge that collapsed yesterday in Minneapolis. And by that I mean real news about the bridge itself as opposed to the people who were unlucky enough to be using it at the time it catastrophically failed.
From ENR.com (Engineering News Record) here's an excerpt from today's report by Tudor Hampton and Aileen Cho, with Tom Sawyer and Tom Ichniowski:
There had been no evidence of additional or growing cracks in welded connections in inspections conducted between 2004 and this year, say Minnesota Dept. of Transportation officials. Based on a study by URS Corp., MinnDOT began a thorough inspection of welded connections inside the two steel arch deck trusses this year, Dan Dorgan, MinnDOT director of bridges and structures, said at a press conference. The inspections began in May and were to be completed after Progressive Contractors, Inc., finished its $9–million contract in September.
St. Michael, Minn.–based PCI, which was performing routine resurfacing work, was still missing 1 of 18 workers as of late Thursday. "Our bridge repair crew of 18 workers was performing routine resurfacing work on the Interstate 35W bridge across the Mississippi River," said PCI bridge division vice president Tom Sloan in a statement. "They were preparing to pour the final two inches of concrete surface. Before they could start pouring, the bridge collapsed. One of our workers has not been located. Three were hospitalized, and others were treated at the scene. All of the survivors are severely shaken up."
He added: "This was a steel structure that supported a concrete deck. We did not do any work on the steel structure. As we have done successfully many times all over the Midwest, we simply repaired the concrete deck by removing deteriorated concrete, patching, and resurfacing."
This bridge is unique because it was constructed with a single 458 ft steel arch to avoid putting piers in the water which would impede river navigation, according to the U.S. Dept. of Transportation.
The bridge was built between 1964 and 1967 by Industrial Construction, says Dorgan. The bridge, designed in-house, consists of steel multi–beam approach spans, concrete slab approach spans and steel deck truss main spans. According to a 2001 report by the University of Minnesota that examined fatigue cracking, the twin steel deck trusses comprise a latticework of riveted plates and I–beams for diagonal and vertical members. "The truss members have numerous poor welding details. Recent inspection reports have noted corrosion at the floor beam and sway brace connection, and rust forming between connection plates," said the report.
The investigators found that peak stress ranges were less than the fatigue thresholds at all details in both controlled and open traffic tests. If fatigue problems were to develop due to increased loads later, cracks would first show up in a floor truss, where they "should be readily detectable, floor trusses are easy to inspect from the catwalk." If cracks went undetected, it added, the bridge could most likely tolerate the loss of a floor truss without collapse, "whereas the failure of one of the two main trusses would be more critical."
The report recommended that the main truss members it identified as having the highest stress ranges be inspected thoroughly every two years. It also recommended inspections of reinforcing welds on lower chords and diagonals of all the floor trusses. "Since they can be inspected easily from the catwalk, they could be inspected every six months," the report concluded.
Speaking on a general level, Stuart Sokoloff, principal with Construction Technology Services, Garden City, N.Y., which provides forensic engineering and failure analysis, noted that "most often when there's a collapse it has to do with the connections and not the actual steel girder." He also notes that "welded connection failures tend to be catastrophic," but are not as easy to find in a visual inspection. "If there is direct shear force on a bolted connection, you'd start to see a prying deformation. A visual hands on would identify it. For a welded connection, unless it underwent some sort of failure, you can't tell what strength is left."
In a hypothetical case, repair work could induced lateral forces that push a truss out of line, he notes. "Trusses are meant to be 2–dimensional entities. The deeper you go from the chords, the stronger the truss you get." However, something like a jackhammer or simple resurfacing doesn't seem likely to cause such a catastrophic failure, he adds.
Richard Stehly, an expert in bridge engineering and co–founder of St. Paul, Minn.–based American Engineering Testing notes that PCI crews were doing maintenance work on the bridge – – patching, partial depth replacement, milling and overlay. " It's hard to see how a partial–depth replacement would be a cause," he says, though nothing can be ruled out.
The critical problem, as noted in the 2001 report, is the lack of redundancy in the truss design. "Two planes of truss support the eight lanes of traffic. If you lose one you lose the whole thing," he says. "The design was fairly common" 40 years ago, he says. "Today you wouldn't build a truss like that."
The possibility has been raised of a train causing vibrations on a track adjacent to piers. While not ruling that out either, Stehly notes that it is a track siding, where trains aren't likely to be going more than 10 mph. But typically, "there's no one factor" in a situation like this.
Which is to say: that the bridge was old, using a design that did not incorporate redundancy. That the failure is most likely to have happened at the connections - the weld joints - and those are impossible to visually inspect for fatigue. That the traffic on the bridge hasn't ever been measured anywhere near calculated failure loads. That the crews working on the bridge at the time were unlikely to have contributed to its collapse, and they're as shaken up by the experience as anyone else who was there. And that it will take some time to figure out just what caused the catastrophic failure of the span.
But hey, the NTSB is on it. They're the guys who reassemble every single airplane that crashes to figure out what went wrong - whether it's a two-seater Cessna or a 747. If they can do that, I bet they can figure out how to straighten a simple steel truss bridge. And maybe what they learn will help us to learn to take care of our aging infrastructure sooner rather than later.
