He's neither a Nostradamus of engineering failure nor a psychic service you can dial for bridge-safety predictions, but Henry Petroski may be the closest thing we've got. Petroski is a professor of civil engineering at Duke University and design guru who has extensively studied—and sung the praises of—spectacular design failures, saying they lead to major progress. This was the subject of his book Success through Failure.
He's also known for popularizing the theory that there's major bridge collapse every 30 years; extrapolating from the West Gate Bridge disaster in 1970, the world was due for another major collapse in 2000. It seems we may have gotten it only seven years late. He theorizes that bridge collapses happen approximately every 30 years because that's how long it takes a new generation of engineers to emerge and then ignore the old lessons, to disastrous results.
The theory first appeared in a 1977 paper in the Proceedings of the Institute of Civil Engineers by civil engineers Paul Sibly and Alastair Walker. They based their theory on observations of the pattern of major bridge collapses: the Dee in 1847 (pdf), the Tay in 1879, the Quebec City in 1907, the Tacoma Narrows in 1940, and the West Gate in 1970.
Petroski, ensconced in a cozy vacation cabin in New England when reached by Discover, hasn’t studied the Minneapolis bridge in detail, but he offered his general insights on such collapses:
Is the Minneapolis Bridge Collapse the disaster predicted by your theory?
The original idea for that every-thirty-years theory comes from [Sibly and Walker], but I’ve sort of projected from that. There was supposed to be one around 2000, but 2007 might still be in the range. A major bridge collapse is something like what happened in Minneapolis, a large bridge that may involve a loss of life, but not necessarily. No one died in the Tacoma Bridge collapse.
Why do bridges like this collapse?
I haven’t really seen pictures of the bridge up in Minnesota, but it doesn’t seem to be a very uncommon bridge for the time it was built. Failures like this tend to happen, not to new daring structures, but to structures that are considered standard. When you’re trying to do something really new you tend to take a lot of care. Building these bridges is supposed to be just careful calculations. The design’s already been done. And also when these sort of standard bridges are built it’s not the really high-level engineers that are doing the detail work on them.
But don’t engineers, like all scientists, learn from their mistakes? It appears that what causes the cycle is that engineers tend to get a little complacent. Engineers are human beings. It’s part of human nature. As you get familiar with something and you think you’ve mastered it, you tend to get a little less careful. It happens to all of us. We’re driving down the road at 70 miles per hour, and you think, “Well, I’ll just lean over to pick something up off the floor,” and there’s an accident.
Where can error creep into the equations? Mostly the analytical tools that are used and the way the stresses and strains are calculated are really old. We may use computers that seem to be doing things better, but they’re really only doing the old stuff faster. There’s a perspective that you’re using computers and they're doing a better job, but that’s an illusion. Computers alone are not necessarily building better bridges.
Is that why you say that in design projects, failure can be a blessing? You think of the Titanic, that it was supposed to be unsinkable. People weren’t vigilant because they thought they had something that could really survive the North Sea. But what if the Titanic hadn’t sunk? People would think, “Well, this is a great design,” and other ocean liners would have been built even bigger. People may have said, “We don’t even need lifeboats, because it’s unsinkable.”
So if something like that is lucky and dodges the iceberg, then other engineers start accepting the design as safe?
There’s a lot of disasters that don’t happen that could happen. When something succeeds, it’s not perceived as luck, it’s perceived as good design. There was really an equally good chance that the Titanic could have not failed.
What are the benefits that can come out of a terrible disaster like this?
If they discover that the bridge collapse was due to metal fatigue, as they seem to be thinking, there will be a renewed effort to look for metal fatigue in other bridges. That’s common sense. This bridge failure will really be a wake-up call. There will be a lot of increased attention to inspection.
What advances in bridge engineering might prevent future collapses? There are new materials. New efforts to inhibit corrosion and new methods of construction. There’s better thinking. And better respect for the forces of nature and the nature of material.
