Forensic Engineering And The Press

This article from the New York Times on the collapse of the central bridge of the highway viaduct in Genoa is worth reading. It covers the topic well and has excellent graphics to help explain some of the more difficult engineering topics. I generally do not want to spend a lot of time discussing current-day failures where I’m simply reading articles and regurgitating them here: in those cases, my input is of little value. But this article got me thinking about the way the investigation of engineering failures is presented to the public.

First, it it’s important to note that the majority of failures that people hear about are not structural. The investigation of plane crashes or car accidents is typically the investigation of mechanical failure. The failures that most people experience often are computer related: hardware or software failures. Structures have a lower failure rate because structural engineering isn’t just a mature field, it’s been a mature field for a long time. Barring an extraordinary event – a hurricane, an earthquake – people expect buildings and bridges to not collapse.

There are a handful of basic ideas that tend to crop up when structural engineers perform forensic investigations. None of them is absolute, but they are useful starting points: Failure happens at connections, not because members are overstressed. No one mistake or problem causes collapse, but rather the combination of several often-unrelated problems does. There were warning signs. Weathering exacerbates bad details. As it happens, all of these were true in Genoa.

There are very few failures that are truly accidents, in the sense of unforeseeable, random events. The vast majority of failures trace back to specific human actions: failure in design, failure in detailing, failure in construction, failure in construction inspection, failure in maintenance, and so on. This is good from the perspective of understanding what went wrong, but in some ways bad from the perspective of learning from the mistakes. People are defensive; people are afraid of being sued for slander. There’s a strong sense that it could happen to any of us and we would want others to go easy on us. If failures were all accidents, none of us need ever be concerned about our jobs or our reputations.

I’ve read a lot of newspapers accounts of structural failure, and they tend toward the dramatic. Descriptions of the horrific failure of the skybridges at the Kansas City Hyatt Regency make for better reading than a discussion of the causes of the failure: a decision to use paired channels welded toe-to-toe as beams, a change to the hanger detail (that accidentally doubled the shear stress in the beams) during construction, an inadequate review of the change. Not every momentous change is dramatic, but a description of engineers sitting at their desks looking at drawings is boring journalism. I’ve been interviewed by reporters and they have a tendency to push for definitive statements, without the qualifiers that come naturally to me. This is, at least in part, toward the good purpose of making a technical topic clear to a popular audience. But if engineers, as a profession, want people to understand why failures occur, we need to get as much nuance into the discussion as we can.