In June 1978, a month after LeMessurier was told of the switch from welds to bolts in the Citicorp building, he received a telephone call from a student. This student's professor had been studying LeMessurier's Citicorp design and had concluded that LeMessurier had put the building's nine-story supports in the wrong place. The supports belonged on the tower's corners, according to this professor, not at the tower's midpoints.
The professor had not understood the design problem that had been faced, so LeMessurier explained his entire line of reasoning for putting the tower's supports at the building's midpoints. He added that his unique design, including the supports and the diagonal-brace system, made the building particularly resistant to quartering, or diagonal, winds -- that is, winds coming on the diagonal and so hitting two sides of the building simultaneously. Pictured is a diagram of why perpendicular winds cause sway in a building.
Shortly thereafter, LeMessurier decided that the subject of the Citicorp tower and quartering winds would make an interesting topic for the structural engineering class he taught at Harvard. Since at the time the requirements of the New York building code, like all other building codes, had covered only perpendicular winds, LeMessurier did not know how his design would fare in quartering winds.
Interested to see if the building's diagonal braces would be as strong in quartering winds as they had been calculated to be in perpendicular winds, LeMessurier did some computations. He found that for a given quartering wind, stresses in half of a certain number of structural members increased by 40 percent.
Then he became concerned about the substitution of bolts for welds. Had the New York contractors taken quartering winds into account when they replaced the welds with bolts? Had they used the right number of bolts? The second question was particularly important -- a 40 percent increase in stress on certain structural members resulted in a 160 percent increase of stress on the building's joints, so it was vital that the correct number of bolts be used to ensure that each joint was the proper strength.
What he found out was disturbing. The New York firm had disregarded quartering winds when they substituted bolted joints for welded ones. Furthermore, the contractors had interpreted the New York building code in such a way as to exempt many of the tower's diagonal braces from loadbearing calculations, so they had used far too few bolts.
Shaken, LeMessurier reviewed old wind-tunnel tests of the building's design against his new quartering-wind calculations (these tests had modeled a large part of midtown Manhattan), and found that under adverse weather conditions, the tower's bracing system would be put under even further stress. The innovative tuned-mass damper, designed to reduce the building's normal slight swaying, was not designed to keep the building from being blown down in a major storm; this further worried LeMessurier.