Within this seemingly simple computation, however, lurks a powerful multiplier. At any given level of the building, the compression figure remains constant; the wind may blow harder, but the structure doesn’t get any heavier. Thus, immense leverage can result from higher wind forces. In the Citicorp tower, the forty-per-cent increase in tension produced by a quartering wind became a hundred-and-sixty-per-cent increase on the building’s bolts.
Precisely because of that leverage, a margin of safety is built into the standard formulas for calculating how strong a joint must be; these formulas are contained in an American Institute of Steel Construction specification that deals with joints in structural columns. What LeMessurier found in New York, however, was that the people on his team had disregarded the standard. They had chosen to define the diagonal wind braces not as columns but as trusses, which are exempt from the safety factor. As a result, the bolts holding the joints together were perilously few. “By then,” LeMessurier says, “I was getting pretty shaky.”
He later detailed these mistakes in a thirty-page document called “Project serene”; the acronym, both rueful and apt, stands for “Special Engineering Review of Events Nobody Envisioned.” What emerges from this document, which has been confidential until now, and from interviews with LeMessurier and other principals in the events, is not malfeasance, or even negligence, but a series of miscalculations that flowed from a specific mind-set. In the case of the Citicorp tower, the first event that nobody envisioned had taken place when LeMessurier sketched, on a restaurant napkin, a bracing system with an inherent sensitivity to quartering winds. None of his associates identified this as a problem, let alone understood that they were compounding it with their fuzzy semantics. In the stiff, angular language of “Project serene,” “consideration of wind from non-perpendicular directions on ordinary rectangular buildings is generally not discussed in the literature or in the classroom.”
LeMessurier tried to take comfort from another element of Citicorp’s advanced design: the building’s tuned mass damper. This machine, built at his behest and perched where the bells would have been if the Citicorp tower had been a cathedral, was essentially a four-hundred-and-ten-ton block of concrete, attached to huge springs and floating on a film of oil. When the building swayed, the block’s inertia worked to damp the movement and calm tenants’ queasy stomachs. Reducing sway was of special importance, because the Citicorp tower was an unusually lightweight building; the twenty-five thousand tons of steel in its skeleton contrasted with the Empire State Building’s sixty-thousand-ton superstructure. Yet the damper, the first of its kind in a large building, was never meant to be a safety device. At best, the machine might reduce the danger, not dispel it.
Before making a final judgment on how dangerous the bolted joints were, LeMessurier turned to a Canadian engineer named Alan Davenport, the director of the Boundary Layer Wind Tunnel Laboratory, at the University of Western Ontario, and a world authority on the behavior of buildings in high winds. During the Citicorp tower’s design, Davenport had run extensive tests on scale models of the structure. Now LeMessurier asked him and his deputy to retrieve the relevant files and magnetic tapes. “If we were going to think about such things as the possibility of failure,” LeMessurier says—the word “failure” being a euphemism for the Citicorp tower’s falling down—“we would think about it in terms of the best knowledge that the state of the art can produce, which is what these guys could provide for me.”
On July 26th, he flew to London, Ontario, and met with Davenport. Presenting his new calculations, LeMessurier asked the Canadians to evaluate them in the light of the original data. “And you have to tell me the truth,” he added. “Don’t go easy if it doesn’t come out the right way.”
It didn’t, and they didn’t. The tale told by the wind-tunnel experts was more alarming than LeMessurier had expected. His assumption of a forty-per-cent increase in stress from diagonal winds was theoretically correct, but it could go higher in the real world, when storms lashed at the building and set it vibrating like a tuning fork. “Oh, my God,” he thought, “now we’ve got that on top of an error from the bolts being underdesigned.” Refining their data further, the Canadians teased out wind-tunnel forces for each structural member in the building, with and without the tuned mass damper in operation; it remained for LeMessurier to interpret the numbers’ meaning.
First, he went to Cambridge, where he talked to a trusted associate, and then he called his wife at their summerhouse in Maine. “Dorothy knew what I was up to,” he says. “I told her, ‘I think we’ve got a problem here, and I’m going to sit down and try to think about it.’ ” On July 28th, he drove to the northern shore of Sebago Lake, took an outboard motorboat a quarter of a mile across the water to his house on a twelve-acre private island, and worked through the wind-tunnel numbers, joint by joint and floor by floor.
The weakest joint, he discovered, was at the building’s thirtieth floor; if that one gave way, catastrophic failure of the whole structure would follow. Next, he took New York City weather records provided by Alan Davenport and calculated the probability of a storm severe enough to tear that joint apart. His figures told him that such an event had a statistical probability of occurring as often as once every sixteen years—what meteorologists call a sixteen-year storm.
“That was very low, awesomely low,” LeMessurier said, his voice hushed as if the horror of discovery were still fresh. “To put it another way, there was one chance in sixteen in any year, including that one.” When the steadying influence of the tuned mass damper was factored in, the probability dwindled to one in fifty-five—a fifty-five-year storm. But the machine required electric current, which might fail as soon as a major storm hit.
... continue reading