The Design Flaw That Almost Wiped Out an NYC Skyscraper 183
Hugh Pickens DOT Com (2995471) writes "Joel Werner writes in Slate that when Citicorp Center was built in 1977 it was, at 59 stories, the seventh-tallest building in the world but no one figured out until after it was built that although the chief structural engineer, William LeMessurier, had properly accounted for perpendicular winds, the building was particularly vulnerable to quartering winds — in part due to cost-saving changes made to the original plan by the contractor. "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," writes Werner. "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 hit New York every 16 years." In other words, for every year Citicorp Center was standing, there was about a 1-in-16 chance that it would collapse." (Read on for more.)
Pickens continues:
"LeMessurier and his team worked with Citicorp to coordinate emergency repairs. With the help of the NYPD, they worked out an evacuation plan spanning a 10-block radius. They had 2,500 Red Cross volunteers on standby, and three different weather services employed 24/7 to keep an eye on potential windstorms. Work began immediately, and continued around the clock for three months. Welders worked all night and quit at daybreak, just as the building occupants returned to work. But all of this happened in secret, even as Hurricane Ella, the strongest hurricane on record in Canadian waters, was racing up the eastern seaboard. The hurricane became stationary for about 24 hours, and later turned to the northeast away from the coast. Hurricane Ella never made landfall. And so the public—including the building's occupants—were never notified.Until his death in 2007, LeMessurier talked about the summer of 1978 to his classes at Harvard. The tale, as he told it, is by turns painful, self-deprecating, and self-dramatizing--an engineer who did the right thing. But it also speaks to the larger question of how professional people should behave. "You have a social obligation," LeMessurier reminded his students. "In return for getting a license and being regarded with respect, you're supposed to be self-sacrificing and look beyond the interests of yourself and your client to society as a whole.""
That has happened quite often here in the US. (Score:5, Interesting)
I wish I could find an appropriate citation - the example I recall was a bridge which needed to be torn apart and repaired because of the use of a different type of bolt securing the framework. The replacement had similar tensile and shearing strength, but several years later the bolts started failing at a much higher than expected rate, requiring the bridge to be retrofitted with the original fastener. It turned out that the new bolt (while actually stronger in some respects than originally required) was subject to vibration stresses. The review permitting the substitution focused on the strength of the bolt required for the application, but the data showing that the bolt was subject to metal fatigue if subjected to extended vibration wasn't available or considered at that time.
Changes such as these are actually not too rare; I suspect that in most cases, the substitutions work exactly as expected, but when we're discussing infrastructure elements of this scope a single failure is not merely troublesome but often catastrophic.
Kansas City Hyatt Regency Skywalk (Score:5, Interesting)
I've heard news reporting before on this subject. The way it goes is this: the architect submits his designs, which are subject to review. Once the green light's given, construction begins. Now, engineers on the project notice a way that they can cut costs or construction time, or somebody requests a modification to the original design (perhaps to add a restroom or breakroom, perhaps to add or remove a wall or subdivide a floor differently). ...
I wish I could find an appropriate citation ...
The Kansas City Hyatt Regency Skywalk disaster [wikipedia.org], 17 July 17 1981, is an excellent case study. Before the collapse of the WTC South Tower it was the deadiest structural collapse in U.S. histories (dam failures are another story entirely). Until 9-11 the CitiCorp Center was well placed to beat it.
In the Hyatt Regency case the design of the double skywalk was changed during constructution, replacing a continuous steel rod that supported both skywalks with two rods, one from the roof to the upper skywalk, and one from the upper skywalk to the lower. Problem was the design had the continuous rod bearing the full load, the change made the upper skywalk bear the load of the lower skywalk (and the people on it) when it was only supposed to be holding up people on the upper skywalk and nothing else.
As built the skywalk was so overloaded that eventual collapse was possible even without any load. Naturally when it did fail it would be at a time when both the upper and lower skywalks were heavily loaded with people, and the floor crowded below. 114 died, 216 were injured - many seriously.
Re:yes, I've used a Professional Engineer. also a (Score:4, Interesting)
What jurisdiction do you live in that actually licenses software engineers?
Re:Hyatt Regency Walkway Collapse (Score:5, Interesting)
The design is fine if you can magically materialize the rods, retaining nuts, and walkways in place, as they appeared on paper. But it's one of those designs where it's completely impractical to get from the disassembled parts to the completed design. The contractor correctly called out this idiotic design and suggested splitting the rods in half - one for the upper walkway, the other for the lower walkway. That way they could connect the rods to the upper walkway, lift it in place and mount it to the ceiling. Then attach the rods to the lower walkway, lift it in place to mount it to the upper walkway.
It was the architect/engineers who didn't properly vet the change. If the two rods had been above/below each other with a mating connector (emulating the original single-rod design), all would have been fine. But the contractor had suggested offsetting the two rods sideways so they could both be sent through the upper walkway, using the walkway itself as the mating connector. That offset (1) transferred the entire load of the lower walkway onto the upper walkway instead of just the rods, and (2) converted what was supposed to be entirely axial loads on the rods into a torque on the walkway floor; a floor whose structure wasn't designed to withstand that much torque, and didn't on the night of the disaster. The engineers should have caught that and come up with a different design.