IN ACCORDANCE WITH ITS NATURE

Reality, to be commanded, must be obeyed

Michael Round

July 17, 2009

It was the summer / fall of 1978 when Hurricane Ella bore down on Manhattan.  As winds ripped through the maze of skyscrapers, a vulnerable 59-story colossus sagged and gave way.  A catastrophic collapse with thousands dead.

You don't remember it?

It never happened.

Why it did not happen is an epic story of heroism!

The structural engineer of the marvelous 59-story Citicorp Center, William LeMessurier, received a call in June of 1978 from a student regarding stresses on the unique skeleton of the 1-year-old structure.  The skeleton was unique.  Supporting columns were centered on each side of the building, rather than on the corners, giving the cantilevered building the appearance of a man on stilts.

The student claimed his professor had told him the structure was not safe, due to stress loads on the odd-shaped building from the Manhattan winds.

LeMessurier assured the student the calculations were right.

The calculations had assumed wind striking the structure perpendicularly.  But a thought came to him, given the positions of the support columns.  What happens if the wind struck the building at a 45 degree angle instead?  Surprisingly, he found the building much weaker than he believed.

It's structural integrity was still much greater than code requirements.  It just wasn't as strong as LeMessurier had thought.

But something else nagged at him.  At a recent and unrelated meeting, the issue of welding joints versus bolting joints had come up.  Bolted joints was the preferred method of construction, as it was cheaper than welding while affording similar strength.  If bolting does not jeopardize the structural integrity of the building, there is no need to over-build a building - or to over-price it.

His initial skeleton structure had called for welded joints, but he had allowed for bolting to be substituted.  This was not seen as a compromise, but instead recognition bolting provided sufficient structural strength cheaply.  All was well.

But these two issues together - the wind-direction assumption and the bolted joints - made the building much much weaker than believed.  The dire predictions suggested the building would suffer catastrophic failure once every 16 years.  16 years! 

Now, at this point, only he knew this.  Nobody else in the world had access to all of this information.  Only him.  What should he do?  What would you do? 

Before you boldly say you'd announce the findings, consider the professional humiliation you'd incur.  Think of the additional costs you and your firm would incur fixing the problem. 

For the whole story on what he actually did do, see here:

http://www.duke.edu/~hpgavin/ce131/citicorp1.htm

An Extension to the Story

There is a related entry here regarding the extension of principles into domains outside the "experience area".  The Tacoma-Narrows Bridge collapsed, we know now, because of the use of plate girders to support the road bed, versus traditional open lattice beam trusses.  A small change leads to a tremendously different outcome.  So too with the Citicorp building, in the extension of principles outside the "experience area".

Fortunately, the issue was caught in time.

More fortunately, LeMessurier was the man to catch it, because of what he did with the information.  Lesser men would have done less.

This Day in History

July 17, 1981

28 years ago.  That's a long time.  It's like yesterday if a loved-one died.  28 years ago, 114 people died as one walkway collapsed onto another at the Hyatt Regency in Kansas City. 

Dancing to "Satin Doll" during a Tea Dance, the 4th-floor walkway suddenly gave way, collapsing onto the 2nd-floor walkway directly below it, both falling to the floor, where even more people were dancing.

Here is the Hyatt the day after:

The intact hanger rods give a clue to how this happened.  The initial walkway design had the 2nd and 4th floor walkway crossbeams suspended from these hanger rods.  The rods were threaded. 

When it came time to actually implement the design, workers protested it was not possible to do this.  A design-change was suggested - and signed off on, where instead of both floors being attached to the rods, the top floor was to be attached to the rod, and the 2nd-floor rod suspended from the cross-beam supporting the 4th-floor.

It doesn't sound like much of a change.  After all, the same amount of weight is still being supported.  Is there any difference?

Henry Petroski, in his great book To Engineer is Human, provides a great analogy.  The design-change was akin to going from two people hanging on a rope to one person hanging on the rope - and the second person hanging onto the first person!

You can immediately see the difference - you can feel the difference!

Most buildings do not collapse.  Most bridges do not fail.  Most airplanes land safely.  99.999%, likely.  But that other .001% is brutal.  We hear about these.  Fortunately, engineers also learn from these.

It's a profession we often take for granted.  We cross a bridge, take the express-elevator to an upper floor of a skyscraper, or land safely often without a moment's notice to how it all happened. 

Perhaps we should give a silent "Thanks" to the people who design and build these "miracles".  There are no miracles.  Just genius - acting in accordance with reality.