I happened to stumble upon a fun little essay: The Bug in the Physical Building, and I was intrigued, and it sent me on a bit of a treasure hunt.
The essay is about the building now known as 601 Lexington Avenue, and more specifically about a particular aspect of that building:
From the beginning, the Citigroup Center was an engineering challenge. When planning for the skyscraper began in the early 1970s, the northwest corner of the proposed building site was occupied by St. Peter's Lutheran Church. The church allowed Citicorp to demolish the old church and build the skyscraper under one condition: a new church would have to be built on the same corner, with no connection to the Citicorp building and no columns passing through it, because the church wanted to remain on the site of the new development, near one of the intersections. Architects wondered at the time if this demand was too much and would make the proposal unfeasible.
Structural engineer William LeMessurier set the 59-story tower on four massive columns, 114 feet (35 meters) high, positioned at the center of each side, rather than at the corners. This design allowed the northwest corner of the building to cantilever 72 feet (22 meters) over the new church. To accomplish this, LeMessurier employed a system of stacked load-bearing braces, in the form of inverted chevrons. Each chevron would redirect the massive loads to their center, then downward into the ground through the uniquely positioned columns.
It turns out that this alternate design introduced some subtle behaviors: The Design Flaw That Almost Wiped Out an NYC Skyscraper:
It was an ingenious, cutting edge design. And everything seemed just fine—until, as LeMessurier tells it, he got a phone call.
According to LeMessurier, in 1978 an undergraduate architecture student contacted him with a bold claim about LeMessurier’s building: that Citicorp Center could blow over in the wind.
The student (who has since been lost to history) was studying Citicorp Center and had found that the building was particularly vulnerable to quartering winds (winds that strike the building at its corners). Normally, buildings are strongest at their corners, and it’s the perpendicular winds (winds that strike the building at its faces) that cause the greatest strain. But this was not a normal building.
LeMessurier had accounted for the perpendicular winds, but not the quartering winds. He checked the math and found that the student was right. He compared what velocity winds the building could withstand with weather data and found that a storm strong enough to topple Citicorp Center hits New York City every 55 years.
But that’s only if the tuned mass damper, which keeps the building stable, is running. LeMessurier realized that a major storm could cause a blackout and render the tuned mass damper inoperable. Without the tuned mass damper, LeMessurier calculated that a storm powerful enough to take out the building his New York every 16 years.
The specific issue wasn't so much about the lead-bearing chevrons, or the winds, or the tuned mass damper, but about what seems like a much smaller issue: the bolts. Citigroup Center, Part 2. The disaster that almost happened relates how "original design's welded joints were changed to bolted joints during construction", and a more detailed report at CrossCurrents provides more information:
LeMessurier’s design and the tower’s construction drawings called for five, full- penetration welded joints in each of the eight-story-high diagonal steel members transferring loads from the tower’s corners to the columns at the center of each face. Offering Citicorp a credit of $250,000, the structural steel fabricator proposed substituting bolted joints. The proposal was accepted. Employing the loads at each joint calculated by LeMessurier’s firm, the fabricator designed bolted connections and prepared shop drawings that were then reviewed and approved by the engineers for fabrication and construction. Although less strong than welded joints, the bolted connections were entirely adequate for the designated loads.
But, at any rate, all this is about the bug.
In some ways, a bug is a bug. They happen.
What interested me about the story was: if you discover a bug in a building, how do you "fix" the bug?
Well, let's return to Chris's original essay:
When the error was discovered, a system was set up to make it possible to evacuate parts of New York City on short notice. Meanwhile, every night after the office workers of 601 Lexington Avenue had packed up, an army of welders opened the building up and welded over the bolted joints. They cleared out by morning when the first busy bees arrived back at their office. This was done in total secrecy. The corners are still the weakest part of the building, but the welded-and-bolted joints are now strong enough to bring the building back well above safety margins.
It's a lot easier with software; you can just issue a patch.
The line between software and hardware is blurring. I have a friend who has a beautiful high-end Fuji camera; he was telling me the other day how excited he was about a software update and how it would improve the autofocus speed and accuracy. I have another friend who has a Tesla Roadster; he was talking recently about the new software that was soon to improve his battery efficiency, cruising range, and acceleration.
But sometimes you have to go back in and repair the bolts: $1.1 million approved for plan to keep Bay Bridge bolts safe
Maroney and the seismic panel asked the committee to approve $1.1 million to design a dehumidification system to dry the rods in the foundation, come up with a grout, lubricant or chemical that will protect the rods and the holes they sit in from water in the foundation, conduct more testing to determine the dangers posed by microscopic cracking discovered in the rods, and purchase jacking equipment to test, treat and aid in future repairs to the rods.
It's just what engineers do: try to make the best design you can, build it carefully, test it, monitor it, and, if it should happen, go back and fix the bugs, er, I mean, the bolts.