Throwback Thursday: Proof testing bridges with living subjects

It sounds crazy now. In an era when iron bridges routinely collapsed under the weight of marching soldiers, the builders of a high-profile exhibit hall chose to prove the sturdiness of their elevated iron walkways by . . . marching soldiers across one. And running them across it. Oh, and having 300 construction workers jump up and down on it, in unison, “for some time.”

Why did they think this was a good idea? And what was the result?

To illuminate this historical nugget, we spoke with Henry Petroski, an expert in failure analysis. A professor of civil engineering and history at Duke University, Petroski included the anecdote in his book To Engineer is Human: The Role of Failure in Successful Design.

A cathedral to progress . . . but would it hold up?

To be clear, the exhibit hall in question was quite innovative in many ways. The Crystal Palace (below) housed the first World’s Fair, in London in 1851. Its soaring glass-and-iron design was like nothing ever seen then. (Almost a horizontal skyscraper, it continues to influence design today.) Cutting-edge tools such as circular saws and steam-driven drills were used in the palace’s construction.

crystal_palace_-_interiorBut as impressive as the Crystal Palace appeared, contemporary Britons raised questions about the strength and stability of the temporary structure, especially its elevated walkways. “After all, during this time iron railway bridges were failing at a rate of almost one in four, and suspension bridges were collapsing under marching soldiers,” wrote Petroski. “The safety of the Crystal Palace galleries had yet to be demonstrated.”

And so, “a 24-foot-square section of gallery was constructed just off the floor on four cast-iron girders,” Petroski wrote. Queen Victoria and the press were invited to witness a proof test. The project’s 300 tradesmen and laborers were assembled, along with a company of sappers and miners (the British military’s engineering corps).

Let’s pause here a moment. Petroski has studied design failure for decades, and he has many thoughts on pedestrian walkways. “I think they’re not taken as seriously, perhaps, as they should be, because the load is considered lighter” than vehicles, he told us. “But when you jam people shoulder-to-shoulder on every square foot of a bridge,” as happened, for example, during the 50th anniversary celebration on San Francisco’s Golden Gate Bridge in 1987, “it’s actually far heavier even than bumper-to-bumper [auto] traffic.” In that case, he said, “the bridge visibly sagged in the middle.”

As to the temporary walkways inside the Crystal Palace in 1851, here’s what happened, according to the London Illustrated News:

The first experiment was that of placing a dead load of about 42,000 lb., consisting of 300 of the workmen of the contractors, on the floor and the adjoining approaches.

The second test was that of crowding the men together in the smallest possible space; but in neither case was there any appreciable effect produced in the shape of deflexion. So much for dead weight.

The third experiment—which was that of a moving load of 42,000 lb. in different conditions—consisted in the same party of workmen walking first in regular step, then in irregular step, and afterwards running over the floor, the result of which was equally satisfactory.

The fourth experiment—and that which may be considered the most severe test which could possibly be applied, considering the use to be made of the gallery floors when the Exhibition is opened to the public—was that of packing closely the same load of men, and causing them to jump up and down together for some time: the greatest amount of deflexion was found to be not more than a quarter of an inch at any interval.

The third experiment was then repeated, substituting, however, the Sappers and Miners engaged at the works, for the workmen of Messrs. Fox, Henderson, and Co.; and this last trial, which was quite as satisfactory as the others to all present, is represented in our illustration [below].


Girding for success

So the walkway held up. How had the palace’s structural engineers been so sure they would avoid embarrassment—and their men would avoid injury—in the presence of their queen and London’s reporters?

As it turns out, the engineering team had tested, individually, all cast-iron girders to be used in the palace and walkways, with a brand-new machine invented specifically for the project. “The ordinary means of testing girders, by loading them with weights, would have occupied far too much time,” according to the fair’s official chronicle. A Mr. C.H. Wild devised “an ingenious apparatus” to accomplish the task in a few minutes (per girder).

hydraulic-pressWild’s apparatus (above) built upon technology developed by Joseph Bramah in the 18th century, and prefigured today’s universal testing machines. It was a hydraulic press that used pistons to squeeze girders “precisely at those points, and in the same manner, as the load from the gallery or the roof would do.” Using the press, Paxton’s engineers calculated that the gallery girders would withstand a pressure of 15 tons, while they estimated that the girders would only be subject to a pressure of 7.5 tons.

Making a circus out of it

The proof test with the jumping and the sappers and miners, then, was done largely for the benefit of the press, said Petroski. And if that sounds risky, get a load of this: “Sometimes bridge designers would walk elephants across them,” Petroski said. For example, in 1874, a test elephant lumbered across the Eads Bridge, over the Mississippi at St. Louis. A decade later, the famous Jumbo did the honors at the Brooklyn Bridge. “It was a mixture of publicity and practicality and superstition.”

The more common way to test bridges back then evokes a classic Calvin and Hobbes strip.  “Once the structure was completed,” said Petroski, “very heavy railroad engines or something equivalent would be driven across the bridge to, quote-unquote, ‘prove’ it would handle the load. That’s a long tradition in bridge-building. And in fact, in Eastern Europe, the engineer who designed the bridge would stand under it—sometimes even with his family—to show this was a solid design that he had all the confidence in the world about.”

But lest you think such stunts are wholly outdated, Petroski pointed to the Millennium Bridge in London. The pedestrian bridge across the Thames opened in June 2000. It swayed noticeably, and was closed three days later. After shoring up the bridge, engineers held a successful test with a hundred volunteers walking over it in 2002.

Lock and load

Nevertheless, in 2017, you probably won’t see human (or elephantine) subjects proof-testing a new structure. Today’s engineers test individual girders and other structural elements before assembly and use the results to calculate the maximum load—as the Crystal Palace team did. But if the public demanded further proof after assembly nowadays, engineers would conduct a load test with simulated human weight.

“They often do ‘drop tests’ with elevators,” said Suffolk Northeast Regional Safety Director Martin Leik, “to test that they can handle the weights they are to be loaded with when in full use.” These are “artificial weights,” noted Leik. “No one would even think about using real human beings for something like this now.”

Well, that is a relief. You might even call it a weight off the shoulders.

Final note: want to learn more about failing bridges? Petroski’s latest book, The Road Taken: The History and Future of America’s Infrastructure, was just released in paperback last week. It’s more timely than ever now.

This post was written by Suffolk Construction’s Content Writer Patrick L. Kennedy. If you have questions, Patrick can be reached at You can connect with him on LinkedIn here or follow him on Twitter at @PK_Build_Smart.