How the St. Louis Arch Stands Against All Odds

And so, in 1947, two years after the end of World War II, a St. Louis citizen's committee sponsored an architectural competition to transform the face of the city's riverfront. Any architects who were U.S. citizens could compete for the honor (and the sweet $225,000 cash prize). "It was the intent of the Jefferson National Expansion Memorial Association that the memorial area would become an integral part of the community's life, and revive the beauty and overall impression of the adjacent downtown area," Moore says.

The two-stage competition included an elite jury of nationally recognized architects like George Howe and William Wurster who favored the more modern style of architecture. Influential names in architecture like Walter Gropius, Charles Eames and Kazumi Adachi were among the 172 entrants, but the judges landed on one lucky contender — sort of.

"When the five semi-finalists were chosen in the fall of 1947, the name Saarinen prompted the competition adviser, George Howe, to telegraph Eliel that he was one of them, and that he should work to prepare a revised design for a second round of judging in early 1948," Moore says. "At the Saarinen studio in Michigan, Champagne bottles were uncorked in honor of the triumph. But a few days later, an embarrassed telephone call from Howe revealed that Eliel was not among the semi-finalists. It was his then-unknown son, Eero, who was the winner of a coveted place as a semi-finalist, with his audacious and creative design of a gigantic, soaring stainless steel arch." Like a good sport and a good dad, Eliel popped a second bottle upon learning Eero was the actual recipient of the news.

Eero officially snagged the top honor on Feb. 18, 1948. In a March 7, 1948, article for the St. Louis Post-Dispatch titled "The Arch Began With Pipe Cleaners" (courtesy of the Gateway Arch National Park Archives), Saarinen wrote that his idea for the arch stemmed from a search to emulate the distinct geometric shapes of the existing monuments to Washington, Lincoln and Jefferson.

"We began to imagine some kind of a dome which was much more open than the Jefferson Memorial in Washington," Saarinen wrote. "Maybe it could be a great pierced concrete dome that touched the ground on just three points. . . . We tried it in a very crude way; the only things we could find to make it with were some pipe cleaners. But the three legs did not seem to fit in the plan, so we tried it with two legs, like a big arch. . . . More and more, it began to dawn on us that the arch was really a gateway, and various friends who stooped to look at what we were doing immediately interpreted it as such. Gradually, we named it the 'Gateway to the West.'"

But constructing a giant swooping structure isn't exactly simple. "In designing the Arch, many studies were made to determine the proper curvature," wrote architect Kevin Roche, who worked with Saarinen on the final design and carried the unfinished plans to completion after Saarinen's death in 1961. In a document from the Gateway Arch National Park Archives titled "Designing the Arch," Roche writes that "while the curve was that of a catenary such as would be formed by a chain hung upside down, the development included studying the impact of a weighted catenary which would slightly modify the simple catenary form. Again — many models, studies, section drawings, many chains hung from the ceiling with weights."

The actual structural design of the Gateway Arch came courtesy of the wisdom of engineer Fred Severud, assisted by an associate named Hannskarl Bandel. In another document provided by the Gateway Arch National Park Archives, Severud's son, Fred Jr., wrote of his father, "Although Eero Saarinen designed the look of the Arch, it was my father and his firm that worked out the way in which it could be built. Without the mathematical calculations worked out by the Severud firm, and the use of orthotropic engineering principles (meaning that the steel inner and outer walls of the Arch are the actual structural members, rather than a skeleton of girders with the steel panels hanging on them), there would be no Arch today. After many years of consultations and hard work, construction was finally begun in 1963."

Severud's son painstakingly detailed the specifics of his father's grand plan. "The Arch was built as two cantilever structures, which eventually met at the top," he wrote. "The design had to consider the loadings and structural action at the various stages, while keeping in mind the practicalities of construction. During the initial cantilever stage, post-tensioned concrete was placed between the inner and outer skins, up to the 300-foot (91-meter) level, to provide the necessary strength for the inward curve of the legs. Above 300 feet (91 meters) the post-tensioned concrete was omitted."

Because the structure was too tall to construct with standard cranes, the crew had to use something called a climbing crane, along with the contractors, on each leg of the arch. "At about 500 feet (152 meter) elevation, a temporary horizontal strut truss was placed from leg to leg, which then made the legs act together, an entirely different stress condition," he wrote.

Figuring out the whole arch part of the arch wasn't easy, either. "The attempt to arrive at a curve for the Arch included suspending a rope fastened at its endpoints; the sagging rope formed a catenary curve, sort of an upside down arch," architect Bruce Detmers wrote in another document from The Gateway Arch National Park Archives. "The adding of weights along the length of the suspended rope changed the shape of the curve. A uniform rope resulted in a curve that was too flat on top. The attempts to weight the suspended rope were not helpful in arriving at a shape to satisfy Eero."

Severud's assistant, Bandel, saved the day in a sense, according to Detmers. "Numerical adjustments were made in the formula and plotted," he wrote. "The plotting process and that of making study models resulted in the final design of the Arch. The legs of the Arch are steeper and larger at the base; the top of the Arch smaller and less flat than a pure uniformly weighted centenary, resulting in the fact that the Arch seems to soar." The final step in defining the Arch involved complex mathematical formulas that took hours to calculate. "Our accounting department's mechanically operated machine could multiply tracking numbers to several decimal places," Detmers wrote. "This by today's standards ancient technology was used to calculate the dimensions of the Arch. I was much relieved to learn that Pittsburgh Des Moines Steel, the subcontractor to fabricate the Arch structure, used a computer to check our calculations and found no errors."

The real issue (as if there weren't already so many others to overcome) was figuring out how to leave just enough space for that final connecting piece at the top — any amount of extra space or insufficient space would wreck the whole thing. And there was an added complication, thanks to Mother Nature. "Since the setting of the final piece was to be done during warm weather, a problem arose," Severud's son wrote. "The legs had different exposures to the sun, which caused them to be several feet different in elevation during the day, so all measurements were made at night, when there was no temperature difference. The construction team planned to also place the piece at night, when both sides would line up."

Apparently, the mayor of St. Louis thought this plan was insane. "This was to be one of the most significant happenings in St. Louis history; the setting must be done during the early afternoon!" Severud's son wrote, describing the mayor's incredulity. "This would cause a difference in elevation which couldn't be overcome; what could be done? My father woke up with the solution during the night: use water to cool down the hot side! The fire chief was contacted, and agreed to provide hoses and pumpers to spray the hot side. The result? The final piece was placed during the day, with no problem."

The final issue to contend with: How in the world were the engineers supposed to bring visitors to the top of the landmark? A regular elevator can't exactly navigate a towering curvature. This problem was solved by Dick Bowser, a college dropout who worked at the Montgomery Elevator Company in Moline, Illinois. His solution: a tram that combined the functions of an elevator with ... a Ferris wheel.

"There were many meetings with Dick, who came up with the idea of the capsule which could rotate as it traveled from the underground station up the leg of the Arch to the visitors' platform at the top," Roche wrote. "We constructed many models of the underground station and built a full-size model of the five-person capsule to test with passengers of various sizes. We also built a full-size cross section through the top of the Arch so that the viewing windows could be small enough to not be visible from the outside but large enough to provide the visitor with a spectacular view. Much effort went into this model to ensure that children and adults of all heights could reach the viewing windows comfortably."

And today, 6,700 daily visitors still use Bowser's innovative system to take in the view from the top of the arch. The 630-foot (192-meter) high arch weighs 43,226 tons (39,214 metric tons), including 900 tons (816 metric tons) of stainless steel, wound up costing $13.4 million, and employed several hundred workers (none of whom died in the process, despite predictions of 13 deaths). The landmark continues to be a major national attraction, drawing 3.5 million visitors each year.