What this story leaves out is that the EVP of Citicorp at the time was a MIT-trained scientist (physical metallurgy). So when this crisis bubbled up - there was no hesitation in action since there was a scientist/engineer in top leadership who was able to communicate to the board the severity of the situation.<p>"Together they flew to New York City to confront the executive officers of Citicorp with the dilemma. "I have a real problem for you, sir," LeMessurier said to Citicorp's executive vice-president, John S. Reed. The two men outlined the design flaw and described their proposed solution: to systematically reinforce all 200+ bolted joints by welding two-inch-thick steel plates over them."<p><a href="http://www.damninteresting.com/a-potentially-disastrous-design-error/" rel="nofollow">http://www.damninteresting.com/a-potentially-disastrous-desi...</a><p><a href="http://en.wikipedia.org/wiki/John_S._Reed" rel="nofollow">http://en.wikipedia.org/wiki/John_S._Reed</a>
One point that this article doesn't mention but the video does (starting about here: <a href="http://www.youtube.com/watch?v=TZhgTewKhTQ#t=350" rel="nofollow">http://www.youtube.com/watch?v=TZhgTewKhTQ#t=350</a>) is that the building wasn't built exactly as designed.<p>In particular, the 8-story-high diagonal parts were done in multiple splices that were supposed to be welded together but ended up being bolted together. It sounds like it made things much worse.
Reminded me of the design flaw of the Millennium Bridge in London, where the engineers accounted for all resonance modes but one, the one that can be caused by pedestrians:<p>> Resonant vibrational modes due to vertical loads (such as trains, traffic, pedestrians) and wind loads are well understood in bridge design. In the case of the Millennium Bridge, because the lateral motion caused the pedestrians loading the bridge to directly participate with the bridge, the vibrational modes had not been anticipated by the designers. The crucial point is that when the bridge lurches to one side, the pedestrians must adjust to keep from falling over, and they all do this at exactly the same time.<p><a href="http://en.wikipedia.org/wiki/Millennium_Bridge_(London)" rel="nofollow">http://en.wikipedia.org/wiki/Millennium_Bridge_(London)</a>
"But what I found out at that meeting were that all factors of safety were gone."<p>Many catastrophic "accidents", and I use quotes because they could have been averted had people not cut corners w/o knowing the full context.<p><pre><code> * Chernobyl (after hours test by an untrained crew with an inverted fail safe design)
* http://en.wikipedia.org/wiki/Hyatt_Regency_walkway_collapse (design change in the field, very similar to the citicorp flaw)
* 3 Mile Island (indicator that triggered on switch rather than valve)
* Fukashima (cost cutting on seawall and generator snorkels)
* Ariane-5 (code reuse, dead code)
</code></pre>
If you want to look at good engineering, look at the Brooklyn Bridge[5] and the DC-3[6].<p>Too many people don't design with proper safety factors. You build it, you test it, you test it till it fails and you understand those failures. I would trust another citicorp wouldn't happen because we can do realistic wind model, we can do an earthquake model, an anything model. Maybe we can get to a safety factor of 1 when everything is automated, when everyone has an off-site backup of their own brain but until then. Safety factor 6.<p>[5] <a href="http://www.asce.org/People-and-Projects/Projects/Landmarks/Brooklyn-Bridge/" rel="nofollow">http://www.asce.org/People-and-Projects/Projects/Landmarks/B...</a>
[6] <a href="http://en.wikipedia.org/wiki/Douglas_DC-3" rel="nofollow">http://en.wikipedia.org/wiki/Douglas_DC-3</a>
The original source is at <a href="http://99percentinvisible.org/episode/structural-integrity/" rel="nofollow">http://99percentinvisible.org/episode/structural-integrity/</a>
Reminds me of: <a href="http://en.wikipedia.org/wiki/File:CCTV_Beijing_April_2008.jpg" rel="nofollow">http://en.wikipedia.org/wiki/File:CCTV_Beijing_April_2008.jp...</a><p>I'm sure the CCTV building is safe, but I get a small panic attack just thinking about walking or jumping up and down in that overhanging corner of the building.<p>How did they fix it? The article says they "welded" but doesn't say what was added to increase the strength of the building.
This has always been a fascinating story - I can't believe Slate's just now discovering it. There's a much longer, more detailed account somewhere.
Yeah. I work in this building and it terrifies me everyday. It does sway in the wind. You can hear the walls creek when the winds are high. Besides the complete fear that is instilled in my heart because this building was built in a flawed way it also has an absolute terrible elevator design.<p>The elevators are double-deckers. So the elevator actually has two rooms that move together. In the morning if you want to go to an even floor you have to enter from the basement level and if you want to go to an odd floor you have to enter from the lobby level. This results in the weird sensation that if you are going to a higher floor the elevators often stop but the doors do not open. Probably related to the weird design the elevators break down often and with or without an elevator set out of commission there are high queue times during peak hours. This is a building that could REALLY benefit from an elevator call system where you enter a floor and it assigns you which elevator to get into to. Why they haven't installed a relatively simple fix like this is beyond me.<p>My final gripe about this building is its sheer ugliness IMHO. The steel facade pegs it as one of those buildings that was built in a particularly time. It's not timeless like the Empire State Building or perpetually modern like the Willis Tower or One World Trade center.<p>I guess its only redeeming factor is the very nice public plaza it has both inside and outside, its direct access to the E and 6 trains (super convenient for me), and the fact that it has not fallen down yet.
If only software flaws in large commercial proprietary/closed source systems were subject this kind of discovery and mitigation <i>before</i> things break.
I think there was an episode of Numb3rs that was based on this. Maybe Season 1 episode 4 (Structural Corruption)?<p>It was fun to see the premise turn up in the show and go, oh, I think I know where they got this idea from.
Maybe this is covered in the New Yorker article marklabedz linked to, but I'm curious what their solution to quartering winds was. (They mention welding as a part of the solution but don't go into more detail.)
DRTA, but I listened to the podcast. In it, they say that if the Citicorp tower fell, it would topple nearby skyscrapers like dominoes.<p>Out of morbid curiosity, what's the worst case scenario for something like this? Let's assume terrorists could detonate a bomb large enough to knock over any one skyscraper in the world. Where should they put it to maximize the chain reaction of destruction?
I wish the author had referenced "Why Buildings Fall Down," the book that first made me aware of the Citicorp Building problem: <a href="http://www.amazon.com/Why-Buildings-Fall-Down-Structures/dp/039331152X" rel="nofollow">http://www.amazon.com/Why-Buildings-Fall-Down-Structures/dp/...</a>
Maybe this is a dumb question, but why couldn't they have found another location for the new church? Why did it have to be on the same plot as the Citicorp building?
More interesting reading around "Tuned Mass Dampers" <a href="http://en.wikipedia.org/wiki/Tuned_mass_damper" rel="nofollow">http://en.wikipedia.org/wiki/Tuned_mass_damper</a>
The first Tuned Mass Damper was in the John Hancock Tower.<p><a href="http://en.wikipedia.org/wiki/John_Hancock_Tower" rel="nofollow">http://en.wikipedia.org/wiki/John_Hancock_Tower</a><p>From an engineering standpoint that is a pretty depressing read.