Falling Down

During the lengthy power outage described in last week’s blog, my library was a refuge, at least in the daytime. Due to limited shelf space (room for some 2500 books assuming a 1-inch [2.5cm] average thickness), my rule-of-thumb is to keep a book only if in principle I might re-read it. Books I read once and see no reason to read again even if I were immortal never make it to my shelves. In truth, most of the books that are on my shelves will not be re-read either – maybe not even re-opened – because, like all humans, I also have limited lifetime. However, all of them would be read again were my lifespan longer, and a substantial minority will be reread regardless. A not insignificant minority are revisited repeatedly. The fragility of the electric grid, all too much in evidence last week, led me to revisit one in particular that I hadn’t re-opened in a couple decades: To Engineer is Human: The Role of Failure in Successful Design by Henry Petroski published in 1985. On re-perusal, it was, as I had recalled, an informative read, and one made more relevant than expected by the bridge collapse in Florida much in the news the past few days.

Petroski was inspired to write the book by the Hyatt Regency disaster in Kansas City. Those old enough to remember 1981 may recall that the Hyatt had two-level walkways called “skypers” spanning a soaring lobby. The skypers were suspended from the ceiling by long steel rods much like a suspension bridge. They were unusually crowded on July 17, 1981, when they collapsed, killing 114 people and injuring 200 more. It remains to this day the largest loss of life in US history from pure structural failure, which is to say without an extreme aggravating force such as earthquake or explosion.

The risk of structural failure is an old concern. It is even addressed in the Code of Hammurabi c.1772 BCE:

229. If a builder build a house for a man and do not make its construction firm, and the house which he has built collapse and cause the death of the owner of the house, that builder shall be put to death.

230. If it cause the death of a son of the owner of the house, they shall put to death a son of that builder.

I can imagine the job site of that family building business: “Hey dad, I really think we should reinforce this wall.”

Civil engineering was not a precise science then and it is not today. Although the theoretical strength of materials can be calculated, the actual strength is never anything close. There are always imperfections and impurities in real-world structural elements. All of them are subject to fatigue, and so have a lifespan. Their actual carrying capacities are given values based on tests and real world experience; a large margin of safety is then added, though this occasionally proves inadequate. In the famous case of the Tacoma Narrows Bridge, the unanticipated correspondence of the resonance factor of the bridge with local winds made a hash of all the engineers’ calculations. The structural integrity of any design is always to some degree hypothetical: “The structural soundness of the Brooklyn Bridge only proves to us that it has stood for over a hundred years; that it is standing tomorrow is a matter of probability, albeit high probability, rather than one of certainty.” There are, of course, limits to the margins of safety we build into constructions for reasons of affordability and practicality. We don’t build cars like tanks, for instance, because no one could afford one and mileage would be calculated in gallons per mile.

Petroski describes failures of railroad bridges, buildings, and aircraft. He also describes how each failure led to better designs.

The failure of the Hyatt skypers, by the way, was eventually traced to the steel rods; they were so long in the original design that for practical construction reasons they were pieced instead. The bolted connections were literal weak links. The collapse in Florida has yet to be analyzed properly, but the fact that construction was not yet complete hints that incomplete structural elements might be to blame.

The good news is pure structural failures are rare. That's not much comfort to families of victims, but it is reassuring to the rest of us. Your chance of death in any given year by this cause is one in ten million, or thirty-two people annually on average in the US. This compares with tens of thousands who die on highways in accidents. Even in Babylon, most builders’ sons were pretty safe.

Thumbs up on Petroski’s aging but still relevant book.

Metric – Speed the Collapse