When engineering fails

An expert explains our cultural fascination with design disasters -- and what the recession means for our safety

Published March 25, 2012 9:00PM (EDT)

 The Tacoma Narrows Bridge known as "Galloping Gertie" during its 1940 collapse.
The Tacoma Narrows Bridge known as "Galloping Gertie" during its 1940 collapse.

As long as humans have been building, they’ve been failing too. Society and civilization, from the first irrigation systems to the Brooklyn Bridge, have been designed by a flawed culture. Sometimes, even with today’s technology, design fails. Bridges collapse, ships sink, apartment buildings crumble. As we build even more daring structures, the likelihood of disaster increases, unless we’re willing to learn from past failures instead of focusing only on past success.

In his latest book, “To Forgive Design: Understanding Failure,” Henry Petroski, professor of civil engineering and history at Duke University and author, previously, of 17 books on engineering including “The Evolution of Useful Things,” explores how structural failure is affected by cultural and economical limitations. By critically examining the interdependency of people and machines related to bridge collapses, airplane crashes and space shuttle failures, Petroski discovers that understanding failure is the only way to bring successful design and engineering into the future.

Salon spoke with Petroski over the phone about human error in design tragedies, how the recession is influencing the future of design, and just how long a bridge can last.

It’s been two and a half decades since the publication of your first book, “To Engineer is Human,” which focused on mechanical and structural failures. Your new book, “To Forgive Design,” takes your original analysis a step further, focusing on the interconnectedness of technology and culture.

I wanted to talk this time about larger systems, things that are much more complicated than just a building or a bridge. Especially things that rely to a large extent on human operators that can make mistakes.

Which are the most fascinating design failures for you?

I have been fascinated with historic bridge failures, in part because they tend to be dramatic and in part because they are so revealing. Over the past century and a half or so, there has been a major bridge failure about every 30 years, which is about the length of time of an engineer’s career. The bridge failures, taken collectively, illustrate how a failure can shock us into paying much closer attention to design. But human beings sometimes have short attention spans, and so they soon forget the lessons learned from the last failure and become careless in designing new bridge types. This carelessness leads to another bridge failure, which occurs about 30 years from the last.

The 1940 failure of the Tacoma Narrows Bridge is especially fascinating, because it occurred almost exactly a century after the engineer John Roebling figured out why early 19th-century suspension bridges were being destroyed in the wind. By studying the failures that did occur, Roebling figured out how to design and build suspension bridges that not only stood up to windstorms but also were able to carry railroad trains — something that no one before Roebling could figure out. Over the next century, suspension bridge designers took Roebling’s masterpiece, the Brooklyn Bridge, as a model of success and little by little eliminated the very features that made it work. Eventually, this led to the Tacoma Narrows Bridge, which was so light and flexible that it was brought down by a moderate wind in 1940. Had the bridge designers of the 1930s remembered and heeded Roebling’s lessons learned, the original Tacoma Narrows Bridge might be still standing.

The most famous bridge disaster in recent memory occurred in Minneapolis, in 2007, and it raised a host of questions about the strength of America's bridge infrastructure. How long can a bridge last?

All bridges are designed to have a specific lifetime. Typically, highway bridges have about 50 years. But over in England, they have iron bridges approaching 250 years. In France there are Roman aqueducts that are approaching 2,000 years old. So a bridge can last a very long time if it's built properly in the first place, and then maintained properly. Large suspension bridges are constantly being painted to prevent caustic air from reaching the steel. Maintenance is very important. Some engineers said you can build a bridge to last forever if you maintain it properly. That’s hyperbole, but it gets across the point. Bridges that tend to collapse by surprise are those that aren’t properly maintained or inspected.

Why are we more obsessed with engineering failures than successes?

Failures are much more dramatic than successes, and people like drama. I think this is why automobile races draw such crowds. People expect spectacular crashes, which we tend to find more interesting than cars just racing around the track. The same is true of bridges, buildings, or any structure or machine. Intellectually, we appreciate the achievement of a success, but after a while successes become commonplace and we do not pay very much close attention to them anymore. Failures, on the other hand, are dramatic. They also teach us a lot more than successes. When something succeeds, we learn little more from it than that it did succeed. When something fails, however, we are driven to try to understand why it failed. This leads us to investigate and study the failure until we get to the heart of the matter. The knowledge gained by studying failure enables us to design more successfully the next time.

Why should we as a culture learn to anticipate failure and focus less on success?

If you do everything exactly the same as the successful example you’re trying to follow, you’ll probably be successful, too. But mostly we make slight changes in what we think are improvements in any successful model, and ultimately that’s what leads us to fail. On the other hand, if we anticipate failure and think, "what will happen badly if I do this?" You can eliminate a lot of bad actions by thinking about failure. Chances of success are greater that way.

How does a recession affect our attitudes toward good design?

The recession has caused a lot of projects to be canceled. Tall buildings especially, even in areas like Dubai, which is booming, have either been canceled or put on hold. Designers are not likely to cut corners on design in this kind of environment, because having something fail or go wrong is more harmful to them. If they design something that turns out to be a failure, it’s not only an embarrassment and potential legal liability but it gives the engineer a bad reputation. There have been plenty of examples throughout history of engineers who have designed bridges that collapsed, and that engineer doesn’t get any more commissions.

There was a call to renew American infrastructure during the recession. Do you think this was a missed opportunity? Do you think it’s too late? And what projects do you think are the most important?

It’s not too late. The question is do we have enough money to answer all the calls to invest in infrastructure. I think the last number I saw for the U.S. alone to bring our infrastructure up to levels that experts think it needs would be $2.2 trillion over five years. That’s the kind of money that people are beginning to understand, it’s a lot of money. Right now in Congress there are a lot of debates going on about transportation, which is just part of the infrastructure. Even finding money for the transportation part of infrastructure is not easy. People are starting to look for alternative ways to fund infrastructure other than government. Back in the 19th century, most infrastructure projects were privately financed. People invested in a bridge with the expectation of getting a return on their investment through the toll charge. Some experts are looking at that kind of model again, as a way of improving the infrastructure while not trading public revenue sources. There’s been a lot of talk about high-speed rail in the U.S., but to do that properly would take so much money that nobody is really talking about it seriously. I think roads are such an important part of America. If the roads get deteriorated to the point where it’s going to cause increasingly much to bring them back, it will be a disaster. Roads and bridges, I would put at the top. Not necessarily a single project, but an all encompassing project.

Is design more or less important in our culture now that it has been in the past?

Design has been important throughout civilization, and I think it will always be important. Everything we do is designed whether we’re producing a magazine, a website, or a bridge. Design is really the creative invention that designs everything. You could argue that a society that isn’t as advanced may be more into design than a country that is advanced. We (America) are sort of on automatic pilot right now, we’ve got plenty of design, design ideas and designers, but we don’t have the money to implement it. So if you interpret design in a much broader sense, to try to fix the system, we are in need of that kind of design. Can we come up with new ideas and new ways of keeping our society at a level of quality of life that we’ve become used to and we think is appropriate? That’s going to take some creative ideas, and creative ideas mean creative design.

When a design tragedy happens, how is it decided where the blame ultimately lies?

There’s no simple answer. First of all, it usually depends on how many people might have been killed. If a significant number of people have died or there is a large environmental impact, there usually will be a pretty high level investigation. Sometimes this even takes the form of a presidential commission as it did when the space shuttle exploded. Sometimes these commissions have a very difficult time pinpointing exactly who might be to blame. In part because of the complexity, but also in part because the sequence of what happened and how the accident progressed is sometimes very difficult to pin down.

I have a chapter in the book on the Deep Water Horizon explosion in 2010. Presently, there are a lot of court cases, and that’s part of what they’re trying to establish, who is responsible? Because in that case there were three or four companies involved. There was British Petroleum, which had a lot of the visibility, but also there was Halliburton, who was operating the rig. And then there was the owner of the drilling rig, Transocean. And each of these blames the others because that’s what happens with contentious legal battles. At any given point, it’s not totally clear who is to blame. In part, because the jury is still out — literally. And in some cases, there are ultimately out-of-court settlements, which close the record and the outside world can’t easily get to it.

Historically there have been examples of very dramatic failures where it was concluded that no one was to blame because everyone was doing their job as expected, it was just that the technology wasn’t completely understood. You’d almost have to look at it case by case to find out who is to blame. One of the things I try to do is point out that a knee-jerk reaction, say when a plane goes down, is to say the plane was badly designed. More often than not, that proves not to be as simple an answer as it might seem at first. That’s why I call the book, “To Forgive Design.”

What’s the golden age of American design?

That would depend on whether you’re talking about big bridges that carried the railroads in the late 19th century or the highway system in the middle 20th century; if you’re talking about industrial design, product design that might have been the early 20th century. To me, it would depend on what aspect of design you’re talking about. My book has a lot of examples of bridges, but I also talk about automobile design safety in the 1960s and '70s when interior safety features became focused on, like seat belts. I don’t think there’s a single answer.

How does America’s relationship to design differ from other countries?

In Europe there tends to be much more checking of designs in an independent way. If someone designs a bridge, another company that’s totally unrelated will check the bridge in a very official way. In America, we tend to rely on self checking within the same company. Practices are becoming global, differences that used to exist are becoming fewer and smaller. So the way a bridge is designed in America or Europe or Asia today pretty much is the same, because the same companies are often involved worldwide. An American engineering company might design a bridge in China, or at least be involved as consultants. Though China is growing and becoming increasingly independent, so they can do everything on their own. A smaller country, if they want to build a tall skyscraper they would generally hire a firm from outside the country because that’s where the experience would be. When Malaysia wanted to build the tallest building in the world a couple decades ago, they hired architects and engineers from the West.

By Megan Wood

Megan L. Wood is an editorial fellow at Salon.

MORE FROM Megan Wood

Related Topics ------------------------------------------

American History History