LEARNING FROM FAILURE

"To Forgive Design" chronicles the most infamous engineering failures of our time.

By MATT RIDLEY

To Forgive Design

By Henry Petroski
(Belknap, 410 pages, $27.95) 

In 1907 in Quebec, a cantilever bridge over the St. Lawrence River—with what would have been the world's longest span—collapsed while under construction, killing 75 workers. The accident was caused by human error: The design engineer did not properly calculate the weight of the structure; the resident site engineer did not heed warnings that the steel was overstressed; and the chief consulting engineer delegated the work to junior colleagues. (click below to read more)

As Henry Petroski notes in "To Forgive Design," his fascinating and occasionally unnerving history of engineering failures, the St. Lawrence accident had many consequences, but two in particular stand out. The first was the determination of Canadian engineers to set and maintain higher standards. The engineers reminded themselves of why they had to aim higher by wearing an iron pinkie ring and observing rituals that included reading the Rudyard Kipling poem "Hymn of the Breaking Strain": "So when the buckled girder / Lets down the grinding span / The blame of loss, or murder, / Is laid upon the man. / Not on the stuff—the man!"

The second consequence, Mr. Petroski says, was that after the disaster in Quebec, engineers around the world began preferring to use suspension bridges to cover wide expanses. When in 1939 a span of record length was required to cross the deep water of the Tacoma Narrows in Washington state, a suspension bridge was the obvious choice. But because the area was thinly populated and heavy traffic was not expected, a two-lane bridge with a narrow sidewalk was designed, and the designer chose to use plate girders rather than deep, open trusswork. The result: a bridge that was extremely narrow, light and shallow for its great length.

High-school science classes to this day are shown a film of what happened: Four months after opening, the Tacoma bridge began moving in a high wind—slowly at first, then gradually building to wild undulations—before finally collapsing. My high-school class was taught that the bridge's movements were an example of resonance and that the disaster could not have been predicted. The second part is not strictly true. The Tacoma bridge collapse is more an example of engineering design failure: Something so long, thin and light was almost bound to be too flexible to stand. Although the bridge did resonate, it was responding to its own movement, not the wind's. "If there was resonance," Mr. Petroski writes, "it was complex and existed between the bridge's motion and the vortices produced by that motion itself."

It is good to learn from "To Forgive Design" what became of the Tacoma Narrows crossing. In 1950, a new suspension bridge opened, with four lanes rather than two to reduce the length-to-width ratio and with open trusswork rather than plate girders for a smaller length-to-depth ratio (and for allowing the wind through). So successful was the new bridge in attracting economic growth to the area that tolls rapidly retired its debt, and within 50 years another was needed to handle the increasing traffic. In 2006, the deck sections of the newest Tacoma Narrows bridge were shipped from South Korea. Unluckily somebody forgot to calculate whether the ship carrying the deck sections could get under the old bridge at high tide, and a minor collision resulted. Mr. Petroski remarks: "Murphy's Law has never been repealed."

The engineering failures that Mr. Petroski discusses do not all involve bridges. He tells of the Titanic's inferior iron rivets and the cracks that grew catastrophically from the corners of the square windows of the British-made Comet jetliners in the 1950s (hence the rounded windows of modern planes). But it is bridges that have the largest claim on Mr. Petroski's attention. Bridge after bridge comes tumbling down in "To Forgive Design"—including the Tay railway-bridge disaster in Scotland in 1879, with the loss of 75 lives; the collapse of the Silver Bridge over the Ohio in 1967 (46 lives); and the failure in 2007 of the I-35 bridge over the Mississippi in Minneapolis (13 lives). The causes ranged from unrecognized metal fatigue to simple overloading. The Waldo-Hancock bridge in Maine, completed in 1931, had a narrow escape: The deterioration of its suspension cables was detected just in time a dozen years ago, and the cables were temporarily strengthened while a replacement span—the Penobscot Narrows Bridge—was built alongside it.

After reading this book, one might be tempted never to venture across a bridge again. But of course that would miss Mr. Petroski's goal: to show how engineers learn from failure and improve their designs. The truth is that bridge builders today know more than ever before about matters like measuring metal fatigue and calculating stress and strain.

The 1967 collapse of the Silver Bridge, for instance, showed engineers how dangerous it was to suspend bridges from chain links that could not be inspected easily. The "eye-bar" chain links had been looked at through binoculars. But even a close-up visual inspection might not have been enough to prevent the disaster: Metal-fatigue cracks could lurk inside individual eye bars. As the official report concluded: "The flaw could not have been detected . . . without disassembly of the eye joint." The Silver Bridge disaster, Mr. Petroski writes, stirred a "renewed urgency" in steel-bridge designers to learn more about the "insidious threat" of metal fatigue "to the integrity of their otherwise graceful and faithful structures." But the collapse's most significant legacy was a regime of regular inspection for all bridges—and the use of designs that could be properly inspected.

For those who enjoy reading about girders and trusses, "To Forgive Design" is, yes, riveting. There are times when the author takes for granted that his audience will share his interest in even the most arcane engineering details, and on such occasions his prose loses its sparkle. But he amply shows the wisdom of the proverb that failure is a good teacher. Even a collapsed bridge leads somewhere.

Mr. Ridley, a Wall Street Journal columnist, is the author of "The Rational Optimist: How Prosperity Evolves."