Wednesday, August 08, 2007

Another research idea...

The recent collapse of the I-35 bridge in Minneapolis has me thinking. I found the entire thing morbidly fascinating. As both a civil engineer and a researcher of technical systems the event poses some unique opportunities. By looking at some images, I suspect that the failure mechanism was a compression failure in one of the compressive members of the approach span leading to a catastrophic failure at the main pier.

What I found most interesting is the public response. People are up in arms about the state of the infrastructure and are decrying the safety of the structures. But if we look at the number of bridges in the US (~500,000), the age of the structures, and the relative infrequency of failures, I think that the reliability is actually pretty good. Increasing that reliability is going to be expensive.

A second response is basically at attack on rational design techniques. Stephen Wolfram, for example, has noted that more reliable (or structurally redundant) bridges will likely look less regular than our current designs do. I don't necessarily disagree but I think his analysis ignores some of the real costs of bridge manufacture. Structural strength and reliability are only one constraint on design, and perhaps not even the most important constraint. The real design goals involve an optimization of cost and aesthetic. I'm not sure how Woolfram's models apply to these constraints.

Ultimately, what most interests me is the differential responses of technologists and the popular press. There may be an opportunity to study how these reponses differ. For example, how did the Globe's coverage of the Quebec bridge collapse differ from that of the technical journals? One resource that may be valuable is UBC's pathfinder on engineering disasters.