Adventures In Epic Engineering Failures

I’ve recently finished The Great Courses’ series entitled Epic Engineering Failures and the Lessons They Teach, taught by my favorite instructor, Professor Stephen Ressler.  Unlike my previous review of his course on Greek and Roman Technology, I think for this review, I will just summarize a few of my favorite lessons.

Lecture 10–London’s “Wobbly Bridge”

On June 10, 2000, London’s Millennium Bridge for pedestrians was opened and thousands of people walked across it.  Almost immediately, two of the spans started swaying when mass numbers of people were on it.  And as the spans would sway, the pedestrians would gradually synchronize their walking until most of the people were walking in step with the movement. The video quality is poor, but here is a short video of this happening.

The bridge was closed two days later, while engineers tried to figure out what the heck was happening. The following explanation is for my fellow non-engineers.

To understand this phenomenon, we need to understand vibration.  Vibration is a reciprocal movement…back and forth.  Hertz is the measurement of how many of these movements occur per second.  One back and forth movement per second is described as one hertz.

This becomes relevant when we are discussing resonance, which is an applied force which matches the natural vibration of an object and amplifies it with each repetition.  We see this when a singer sings a certain note and shatters a glass.  The natural vibration of the singer’s voice at a certain pitch (a “high note,” usually) matches the natural vibration of the glass.  And if the singer sustains the note (the applied force), the amplitude (how wide the back and forth of the vibration is) of the glass’ natural vibration expands with each vibration cycle until the vibration exceeds the durability of the glass and it shatters.

An even simpler way to visualize this is to think of a person on a swing.  They are swinging back and forth in a gentle arc, which we can think of as the amplitude of their vibration.  Suddenly, a “helpful” soul stands behind them and every time the person starts to swing forward, the “helper” gives a push.  With each push, the arc gets larger; another way of describing this is that the amplitude is increasing with an applied force.

In the case of London’s Millennium Bridge, which acquired the unfortunate nickname of the Wobbly Bridge, the natural vibration of the bridge spans was about one hertz: that is, one back and forth vibration per second.  This is closely matched by the average walking pace of pedestrians, which is also close to one hertz.  As thousands of pedestrians walked across the bridge, the applied force of their footsteps generated resonance which increased the amplitude of the bridge’s vibrations, causing it to sway laterally.   This problem was increased by the fact that pedestrians walking on a moving surface instinctively match their steps to the movement.  So those thousands of pedestrians were walking together as if they were marching in step, which really concentrated the applied force they were supplying to the structure.

The engineers trying to address the problem did multiple tests, naturally.  One of the tests was having increasing numbers of pedestrians walk across the bridge.  The magic number was 166.  Once that many pedestrians were walking all at once, the bridge started to move, the pedestrians started to fall into step with one another, and the bridge movement increased so dramatically that the test was discontinued.

This discovery actually generated a new name for a previously unidentified phenomenon: synchronous lateral excitation.  You can see this happen in this YouTube video where four metronomes sitting on a single beam are all set to the same beats-per-minute, but started at different times, so they are not moving synchronously.  However, watching them tick for less than a minute, you see them all synchronize.  It’s quite astounding, really.

The solution the engineers finally implemented was tuned mass dampers.  Think of this as shock absorbers for structures.  The dampers would respond to any increased amplitude in the bridge and generate a vibration in the opposite direction.  Of course, this is an extremely simplified explanation.  But if the bridge’s vibration was moving right, the dampers would move left, absorbing and reducing the amplification.  And it worked like a charm.

The bridge has been reopened to pedestrians and fulfills its purpose beautifully.  But it is slow to shake the moniker of “Wobbly Bridge.”  Perhaps someday…

Lecture 18–the Leaning Tower of Pisa

When I was a kid, someone told me that the leaning tower of Pisa was built that way on purpose. It wasn’t falling down…it was just designed to look like it was falling down.  That, of course, is complete nonsense but I hadn’t realized that had stuck with me until I watched this lecture.  Construction of the tower began in 1173 and even while the first story was under construction, the tower started to lean toward the north.  Masons tried to angle their stonework to bring the tower back into true and construction continued for around five years, at which time only three levels had been completed.

Historical and political events halted the tower’s construction for about a hundred years.  When construction began again in 1272 and the tower climbed up to seven stories tall, it suddenly began leaning to the south.  In 1284, construction was halted for another 70 years.  When the belfry was finally built, the builders tried to adjust its position to bring the axis back into true but were unable to compensate for the substantial weight.  Over the next century, the tower’s southern lean went from about 2.5° to 4.5°.

Over the following centuries, various architects, builders, and engineers attempted a number of solutions to stop (or at least slow) the continually increasing lean.  But all of them were either ineffective or else made the problem visibly worse.  By the 1980s, the building’s lean (now 5.5°) was causing obvious structural stress and damage, and the Italian government finally got serious about a solution.

Without going into too much detail, the underlying problem (pardon the pun) was soil mechanics.  And therein also lay the solution.  Engineers under-excavated soil from the north side of the tower.  That is, they used an auger, and over about 18 months, they pulled soil out from under the north side of the foundations one cubic foot at a time, carefully watching the response of the tower.  This has stabilized the tower so effectively that engineers estimate it will remain standing for at least 300 years, barring any unforeseen events (earthquakes, floods, bombs, etc).

Lest we get too cocky about how much more advanced we are than medieval builders, the Millennium Building in San Francisco is undergoing its own increasing lean, again because of the properties of the soil on which it’s built.  The solution to that problem is ongoing.

Lecture 22: The Challenger Disaster

I was in my first year at a Christian college in 1986 when the Challenger shuttle exploded after takeoff.  As soon as the college administration heard what happened, every student was pulled out of their classes and brought into a school-wide chapel service to watch the news and then pray for the families left behind.  Watching the footage, all of us were in a state of stunned disbelief.  How could this happen?  Well, having watched this lecture, I now know how it happened.   I feel this particular tragedy more personally than most others in this course since I was alive and watching when it happened.  And it makes me both sad and angry that there were so many chances to prevent the disaster, all of which were squandered.

The underlying cause of the disaster was faulty O-rings in the solid rocket boosters (SRBs).  The SRB had been designed in the 70s and as far back as 1977, there were known problems with the design, particularly the O-rings.  After multiple problems and studies and experiments, it was clear that under certain conditions, the O-rings could fail catastrophically. But they hadn’t yet.  As  Richard Feynman says, “When playing Russian roulette, the fact that the first shot went off safely is little comfort for the next shot.”

There were multiple delays of the Challenger launch, mostly based on small problems that cropped up and unseasonably cold weather, which was known to make the O-rings more brittle and therefore less effective.  A proposed repair to the O-ring joint was not approved by NASA because of the cost.  NASA was almost certainly getting pressure from above to continue their launches on schedule and put a great deal of pressure on Thiokol (the company that manufactured the SRBs) to agree that the January launch did not need to be postponed further.  I will spare you the back-and-forth, but suffice to say that the Thiokol engineers all agreed that the launch should be delayed.  Their manager, who was an engineer who had moved into administration, told NASA that the O-ring tests were “inconclusive” and therefore the launch could proceed.

This really boggles the mind.  The tests were not inconclusive.  But even if they were, that would be more reason to delay, not go ahead.  But go ahead they did.  And the Challenger exploded, all hands on board lost.

Professor Ressler points out that the usual narrative of “courageous engineers versus corporate stooges” is self-serving.  All of the management personnel involved in the final decision to launch rather than delay were also engineers.  “Being promoted into management did not relieve them of their ethical responsibilities as engineers.”

I really loved this entire series. There were a few lectures where I felt a little bit out of my depth, particularly the lecture about the Deepwater Horizon disaster. That lecture leans heavily into fluid mechanics, and for some reason, I just could not wrap my brain around it. But I still enjoyed the presentation. And I sure did learn a lot. I can’t recommend these courses and this professor highly enough.  Next up: Professor Ressler’s course Everyday Engineering.