On the morning of October 28, 1965, a crew of ironworkers stepped onto a platform at the top of two steel legs rising 630 feet above the banks of the Mississippi River in St. Louis, Missouri. The legs had been climbing since 1963. Each weighed thousands of tons, and after two years of construction they had to meet at the top with a tolerance no greater than 1/64 of an inch, thinner than the cover of a matchbook, at the height of a 63-story building on a bend of the river.
In their hands was a wedge-shaped piece of stainless steel, eight tons of metal, the final section of the Gateway Arch. The gap between the legs had to close. The engineering records of the Jefferson National Expansion Memorial, the name the government gave this project long before its final form was known, tell the story of a structure that demanded almost unprecedented precision on this scale. They also tell the story of a design competition that almost did not happen, and a site cleared of 37 city blocks before the first shovel ever touched the ground.
Two stories twist through the whole thing. The first belongs to a Finnish-American architect who submitted a design so strange that the review committee was not sure what it was looking at. The second belongs to the riverfront itself, the land the Arch stands on, and the people who were actually there. St.
Louis, 1764. Pierre Laclede and Auguste Chouteau chose a limestone bluff above the west bank of the Mississippi as the site of a fur trading post. The location was almost perfect. The Missouri River meets the Mississippi just upstream, and the Ohio joins it as well.
Anything moving through the heart of the continent by water passed that bend in the river. Within decades, St. Louis became a gateway in the literal sense. You passed through it to go west.
By 1850 it was the second-largest port in the country. By the Civil War era, the St. Louis riverfront was packed with commercial offices, warehouses, and trading posts along the levees, dealing in cotton, lead, furs, and grain. The wharves stretched for miles, steamboats stacked in rows of five or six at the docks.
Then the railroads arrived, and they arrived fast. The first bridge across the Mississippi at St. Louis opened in 1874, the Eads Bridge, an iron arch built by James Buchanan Eads against the opposition of most engineers in the country. Once that bridge stood, railroad shipments no longer had to stop at the river and transfer to a boat.
St. Louis, the city that had prospered precisely from that transfer, began to lose its reason for being exactly where it was. The population kept growing for a while because cities have momentum, but the economic logic was fading. By the 1930s, the near north riverfront that had built the city’s wealth was badly deteriorated.
The warehouses that served the steamboat trade sat empty or had been subdivided. The streets along the levee were a maze of old commercial buildings no one had invested serious money in for decades. In the fall of 1933, a lawyer named Luther Ely Smith walked along the riverfront and saw what the real estate market had already decided it did not want. He saw a site.
Smith was a booster, a past president of the St. Louis Bar Association, the sort of man who spent his evenings thinking about what a city should be. What he imagined for those forty blocks along the river was a national monument, not a museum, not an office complex, but a memorial to the westward expansion of the United States, the history that had made St. Louis matter.
Smith put the idea to Mayor Bernard Dickmann, who put it to Washington. On December 21, 1935, President Franklin Roosevelt signed an executive order creating the Jefferson National Expansion Memorial on the St. Louis riverfront. The federal government would acquire the land.
The city and state would contribute matching funds, and somewhere on those thirty-seven blocks, once they were cleared, a great national monument would rise. What that monument would be, no one had yet determined. That decision took twelve years. Before any plans were drawn, long before any competition, the clearing began.
Between 1935 and 1942, the federal government acquired nearly forty blocks of the old St. Louis riverfront, about 82 acres. Thirty-seven of those blocks had buildings on them. The structures removed ranged from nineteenth-century commercial warehouses to small shops, hotels, and apartment buildings.
Most of the commercial buildings were underused, but not all of them. Some were still operating. The businesses in them had leases, customers, employees. The people living in or near the area had to find somewhere else to go.
Federal condemnation law required that property owners be compensated, but it did not require that the compensation be adequate, or that displaced residents have somewhere to go. So the eviction went ahead. By 1942 the land lay barren along the riverfront below the Old Courthouse. Only one building remained, the Old Courthouse itself, the domed building where Dred Scott had argued before the Missouri Supreme Court in 1847.
It stood at the western edge of the memorial site, and it still stands there today. The cleared ground stayed mostly empty for years. The design competition for the monument was not held until 1947. That gap between clearance and design says something about how projects like this actually run.
Land is easier to clear than ideas. A national competition was announced in late 1947. It was open to American architects and asked for a design of the entire memorial site, including the monument itself, the surrounding grounds, and the underground museum that would lie beneath an open park. The jury included some of the most prominent names in American architecture and landscape design.
The first prize was $40,000, and the competition received 172 entries. One of those entries came from Eero Saarinen, a 37-year-old architect practicing in Bloomfield Hills, Michigan. His father, Eliel Saarinen, had been one of the leading architects of the previous generation. Born in Finland, Eliel designed Helsinki Central Station before moving to the United States.
Eliel submitted an entry as well. The two men submitted separately. When the jury began reviewing entries in early 1948, a story began that has been told ever since: that the committee telephoned the elder Saarinen to tell him he had won, only to discover it had reached the wrong Saarinen. The records contain no version of this story that identifies the caller, the date of the call, or any supporting documentation.
What the records do show is that the jury announced a winner on February 18, 1948. The winning design was Eero Saarinen’s, entry number 144. It was a drawing of a structure unlike anything built in America before. Saarinen proposed a stainless steel arch rising 630 feet over the St.
Louis riverfront. It was not a gateway in the literal sense, not a frame to pass through, but a free-standing steel curve, an arch like the mathematical symbol. The shape was not a parabola. Saarinen consulted engineers and chose a weighted catenary, the curve formed by a heavy chain when its links are denser at the ends.
A parabola distributes its weight evenly, but a weighted catenary concentrates load toward the base, giving the legs a sharp slant at the bottom and a flatter arc at the top. The shape actually looks like a curve leaning steeply toward the sky before it reaches the top. The proportions Saarinen set were precise. The Arch is 630 feet tall and 630 feet wide at its base, a perfect square measured from leg to leg and from the ground to its peak.
This number was not chosen primarily for aesthetics but for the structural engineering of a stiffened curve. The Arch solves two problems at once. It transfers load outward through its legs into the ground. The deeper the base, the more the ground can bear the load.
At 630 feet across, the foundations spread far enough so that the pressure from each leg meets the earth at an angle the bedrock beneath the site can withstand. At 630 feet high, the Arch can be seen from most of the city, from the river, and from the Illinois shore opposite. Saarinen’s design was a structural solution and an urban one at the same time. The jury awarded the prize in early 1948.
Saarinen worked with the federal government through the late 1940s and early 1950s to refine the design. The structural calculations were developed with the engineering firm Severud-Elstad-Kruger Associates. Hans Karl Bandel, a German-born engineer working for that firm, took on the main structural work for the Arch. The problem Bandel was solving was essentially an engineering one.
The Arch had to be built as two separate legs rising independently from their foundations. No scaffolding could support a structure this tall on that site, and no temporary framework was practical at that scale. The legs had to be self-supporting at every stage of construction, carrying their own weight as they grew, until the moment the legs met and the Arch became structurally complete. That moment, the closing of the Arch, was the hinge from which every other element hung.
If the legs rose even slightly off their calculated angles, they would not meet at the top. The pieces were converging in midair. Bandel and his team solved this problem in a way that seems from another era. They continuously surveyed the alignment of each leg as it rose.
Every section added to either leg was measured against the calculated geometry of the catenary curve, and any deviation from the expected angle was recorded. The structure was not just built to a plan; it was steered toward a target. As each leg neared the top, engineers fed the measurements into calculations and adjusted tolerances to account for what had happened in the lower sections. The tools available in 1963 included traditional surveying instruments and IBM computing equipment that contractors used to track the cumulative geometry.
For a structure of this kind in that era, that combination was at the edge of the possible. The construction contract went to a joint venture, MacDonald Construction of St. Louis partnered with Pittsburgh-Des Moines Steel of Pittsburgh, usually referred to in the records as PDM. MacDonald handled general construction.
PDM fabricated and erected the steel structure. Site work began on February 12, 1963. That date was Abraham Lincoln’s birthday, and the choice was deliberate. The Jefferson National Expansion Memorial rested on the idea of westward expansion, and Lincoln’s birthday was a symbolic date that government projects of that era were designed around.
The Arch is not solid steel. Each leg is hollow, a triangular tube that narrows as it rises. Each side of the equilateral triangle at the base of each leg is 54 feet long; at the top, each side is 17 feet. The outer skin is stainless steel, quarter-inch thick, polished to the reflective shine seen today.
Inside the outer skin is a second skin of carbon steel, thicker and structural. Between the two skins, in the lower sections of each leg, the space is filled with concrete. As the legs rise, the concrete stops and the interior becomes hollow. The concrete in the lower sections exists for the same reason a tree trunk is wider at its base.
The load is greatest at that point, and concrete adds mass and stiffness where the Arch needs them most. Assembling the Arch required a machine capable of moving along the curve as the legs grew. The contractors used crawler cranes, essentially climbing cranes that moved up the outer surface of each leg on a set of tracks, inching upward as new sections were added. An Arch section was not a ring but a curved equilateral triangle, a piece of triangular tube, roughly 8 to 12 feet long, prefabricated at the PDM plant and shipped to the site.
Each section had to match the one before it perfectly. Tolerances were set in the fabrication plant according to Bandel’s calculations. Any section arriving at the site out of specification would carry its error into every section above it. Two crews worked at once, one on the south leg and one on the north.
In photographs from that period, the cranes look like iron spiders clinging to the sides of two steel triangles leaning slightly toward one another over the St. Louis skyline. Each leg grows in those photos like a column of water, slowly, and from a distance, vertical. Up close, the lean is visible, maybe two or three degrees off vertical at the base, sharpening as the leg rises, the catenary curve asserting itself against the straight line the eye expects.
A labor dispute in the summer of 1964 slowed work on the Arch for several weeks. The ironworkers on the crawler cranes were represented by the International Association of Bridge, Structural, Ornamental, and Reinforcing Iron Workers. A jurisdictional dispute, specifically over which union should operate the cranes on the outer skin of the Arch as it rose, halted work. The dispute was eventually resolved and work resumed.
The names of the men who returned to the cranes after that stoppage are not in the records. The number of men working on each leg of the Arch at any time ranged from twenty to forty-four, depending on the stage of the work. By the spring of 1965, both legs had passed 400 feet. Skyline photos from that period show the Arch in a state that looks more like an unresolved question than a finished idea.
Two leaning steel columns rising at angles so slight they seem impossible to meet. Visitors to the construction site, which drew spectators even while under construction, consistently reported that from the ground the legs looked like they were leaning away from each other, not toward each other. The catenary curve is so subtle at lower heights that the eye reads it as nearly vertical. Only from a distance, and above a certain height, does the curve become clear.
That is worth noting because much of what Saarinen designed was not meant to be understood up close. The Arch is a structure that makes its argument from the opposite bank of the river. The closing of the Arch was scheduled for October 28, 1965. By early October, the two legs were only a few feet apart at the top, and right at the finish line, the weather created a problem no one had anticipated in the schedule.
Steel expands with heat and contracts with cold. The legs of the Arch are made of thousands of tons of steel. On a sunny October day, the south leg, which faced the sun more directly in the morning hours, expanded slightly more than the north leg. The expansion was not large, only inches, but at the top of a structure where the legs had to meet within 1/64 of an inch, inches mattered enormously.
The gap at the top of the Arch was not symmetrical. It changed hour by hour with the sun’s movement and the shadows of clouds passing over one leg. The construction team used hydraulic jacks as the remedy, and the method is described in engineering texts as audacious, a term used in construction engineering to mean the procedure succeeded, but no one was entirely calm while it was carried out. The jacks were placed inside the bases of the legs and used to push them slightly apart.
Because of the shape of the catenary curve, pushing the legs apart at the base caused their tops to close together. By widening the base and waiting for the steel to straighten at the top, the team could insert the final keystone into the gap. A temporary steel frame was also placed across the gap at the top of the structure to hold the legs in position while the final sections were installed and welded. Fire crews stood by with water to cool the south face of the south leg, reducing its temperature and therefore its expansion, while the jacks worked at the base.
On the morning of October 28, the temperature was about 60 degrees Fahrenheit. Wind speeds at 600 feet were higher than at ground level, as always. The crawler cranes were repositioned for the final lift. The keystone, the last section of the Arch, was a wedge of stainless steel made to fit the gap that calculations predicted would exist at the top when all sections were aligned.
It weighed about eight tons. The ironworkers on the platform guided it into place. The hydraulic jacks at the base kept the legs apart. The keystone was inserted.
The final welds were made. The gap closed. The Arch was complete. What happened next is the part of this story that usually gets the least attention, perhaps because it is not glamorous, or perhaps because it is underground.
The Arch is not just an arch. Saarinen’s original competition design included an underground museum, and that museum was built. The Museum of Westward Expansion lies beneath the park at the base of the Arch and opened to the public in 1976. The visitor center and museum complex were redesigned and expanded in a project completed in 2018.
The museum now extends beneath the full footprint of the Arch grounds and is reached through a series of glass structures at the north and south bases of the Arch. But what most visitors remember, and come for, is the tram. Getting to the top of the Arch requires riding a system unique in the world, unlike any other building. The legs of the Arch are hollow, and inside each leg, from base to top, runs a tram line.
These trams are not conventional elevators, because an elevator runs vertically, while the interior of each leg curves. A conventional elevator car rising through a curved shaft would constantly hit the walls. Instead, the designers built a system of small capsule-shaped cars mounted on rotating frames, each capsule seating five people. The capsule rotates around its frame to stay level as the frame moves along the curved track.
At the base, where the leg is nearly vertical, the capsule sits almost flat. As the track curves upward, the frame tilts, but the capsule rotates to compensate. At the top, where the interior shaft is almost horizontal, the capsule has rotated to become vertical to the floor. The capsules are oval steel pods, eight capsules per car, running in a chain along the inner leg.
The trip to the top takes about four minutes. From the observation deck at the summit, visitors look out through small windows set in the stainless steel skin. On a clear day, visibility is 30 miles in any direction. To the east, the Mississippi River, the bridges, and the Illinois floodplain.
To the west, downtown St. Louis and the Old Courthouse. If you look straight down the outer curve of the upper Arch, you see it narrowing gradually to its base below, then the base widening out over the park and the city. The structure is 630 feet tall, and the windows sit at 630 feet too.
You are standing at the top of the weighted catenary, the point where all of Bandel’s calculations converged. The Arch sways. Its design allows a maximum deflection of 18 inches at the top in high winds. During a strong storm, visitors on the observation deck can feel the structure move.
Saarinen set the allowable deflection because any rigid building of this height would collapse under dynamic wind loads before a flexible one would. The Arch bends slightly the way a tall tree bends, and by bending, it stands. The steel does not crack because it is not designed to be unyielding. The Arch was dedicated on May 25, 1968.
Saarinen was not there. Eero Saarinen died of a brain tumor on September 1, 1961, two years before construction began and four years before the keystone was set. He knew his design had been accepted; he had spent years refining it with engineers and the federal government. He reviewed competition models, structural drawings, and site studies.
But he never saw his Arch. The man who designed it never saw it built. What that fact means depends on how you read the story. If you read it as a story about construction, Saarinen’s death is a footnote; the Arch was built to his specifications and it stands.
If you read it as a story about a person and the meaning of a working life, it lands harder. The most celebrated work of architecture associated with Eero Saarinen’s name sits in a city he never visited after its completion. When the Gateway Arch opened to visitors in the early 1970s, the city was suffering what demographers later described as the steepest population decline of any American city in the twentieth century. The city’s population peaked at about 856,000 in 1950, fell below 622,000 by 1970, and below 453,000 by 1980.
The signs of that decline were already visible in 1965 when the keystone was set. The neighborhoods around the memorial site were not the commercial boom area they had been in 1935; they were buildings with empty storefronts and deferred maintenance, with continuous out-migration to the suburbs of St. Louis County. The Arch was built to revitalize the riverfront.
That was part of the argument for it from the beginning. In 1933, Luther Ely Smith argued that a national monument would draw visitors, investment, and civic pride. In the decades after the Gateway Arch opened, some of that came true. Hotels were built on the riverfront.
Gateway Arch National Park, as it was officially renamed in 2018, draws large numbers of visitors, ranking among the most-visited national parks in the country. On summer weekends, the tram lines stretch long. The park at its base is one of the most-used urban parks in the Midwest. But the city itself did not turn around.
The population decline eventually slowed, but it did not stop. By the 2020 census, St. Louis had about 300,000 residents, down from 856,000 in 70 years. The Arch did not fix that.
No monument alone could. What it did was give the riverfront a reason to exist after the steamboats were gone. And the riverfront, at least that 82-acre stretch where Luther Ely Smith walked in 1933, looks better than it has in a hundred years. The fate of the people forced out of that stretch to clear the land for the monument is harder to pin down.
Federal condemnation records list property owners’ names and itemize compensation amounts. They do not record where those people went, whether their compensation was enough to rebuild what they had, or the cost of displacement in terms of businesses and community networks that cannot simply be moved to a new address. The monument the Arch stands on was built by removing something to make room for something else. What was removed was a neighborhood judged blighted by the people who wanted to build there.
Whether the residents and workers of that neighborhood shared that judgment is a question the condemnation proceedings never asked. One specific name from the engineering side of this deserves mention because it is rarely cited in popular accounts. Hans Karl Bandel was born in Fürth, Germany, in 1925. He emigrated to the United States after the war and joined the consulting engineering firm for the Arch’s structural design.
He was in his mid-thirties when the structural calculations for the Arch were made. His design included the closing method, the continuous surveying, the tolerance tracking, and the hydraulic jacks at the base. In accounts that focus on Saarinen’s vision, Bandel is usually mentioned in a single sentence, if at all. In the construction history, he gets a paragraph.
The precision demanded at the top of the Arch was not achieved by stainless steel or the weighted catenary equation alone, but by an engineer who spent years devising a way to build a curve in the air with nothing above to support it. The St. Louis newspapers and national wire services covered the closing ceremony on October 28, 1965. The ironworkers who installed the final keystone are identified in some accounts by their local union and their job roles, but not by their individual names in the records.
They were men who had worked on the crawler cranes for two years, men who knew well the feel of working inside a hollow steel triangle 600 feet above the river in October winds. The keystone they installed weighed eight tons and filled a gap defined by two years of calculations and continuous measurements. When it was set, the Arch became what it was meant to be. The workers came down off the cranes and the Arch stood.
If you want to see it now, how you get there matters. The best view is not from the ground at the base, although that view has its own appeal, the stainless steel surface changing color with the light, the curve so pure it looks more like a drawing than a building. The view the Arch deserves is from the east bank of the Mississippi, from the Illinois side, on the bridge, or from the riverfront park in East St. Louis.
From there, the Arch rises over the water, and you see its full width, 630 feet from tip to tip, against the St. Louis skyline. From that distance the curve is unmistakable. The weighted catenary does its job exactly as Saarinen designed it.
It rises steeply, then leans, then closes. The two ends meet at a point 630 feet up, and the eye follows them to the top without finding the joint. Inside the museum, the exhibits cover the history of westward expansion, the Lewis and Clark expedition, the fur trade, the settlement era. Near the entrance there is material on the design competition and construction.
The structural drawings are there, and the competition design. Photographs from 1963 to 1965 show the two legs rising, the climbing cranes going up, the gap at the top narrowing gradually over months. One photo from late October 1965 shows the legs a few feet apart at the summit, the temporary steel frame in place, the final gap ready. That photo was taken from the work platform 630 feet above the city.
In that photo, the legs are so close that a person standing on that platform could have put a hand in the gap. Within hours, the gap would be gone. Two years of construction, twelve years of design and negotiation, thirty years from Luther Ely Smith’s walk along the riverfront to the day the keystone was set. All of it converging on a wedge of stainless steel eight feet wide at its top.
The tolerance allowed was 1/64 of an inch. That number stays with anyone who studies this. Not because it represents the greatest engineering challenge in American construction history, but because it is the cleanest example of a problem that faces every big project: the gap between what you designed and what you built. On the highway, that gap shows up as sixteen miles of mud in the Texas Panhandle.
On the canal, it shows up as nine million dollars spent before the work was done. The Arch shows it as a gap at the top of two steel legs that rose independently for two years, measured in fractions of an inch. The engineers who solved it used hydraulic jacks and a fire hose, and the fact that steel changes size with temperature. They used what they had, and the gap closed.
From the Illinois shore, looking across the river at the Arch and the city behind it, you see a structure built on a site whose buildings were removed thirty years before construction began, designed by a man who died four years before its completion, and closed by a crew whose names are absent from public records. What it represents, as the government described it, is westward expansion, the idea that what made America was its movement. But in fact, it is a fixed point, made of steel and calculation, that does not move. It stands where the gateway was, 630 feet of curve over the river, and it lets the city change around it.
The Arch Saarinen drew in 1947 is the same Arch standing in St. Louis today. The proportions have not changed, nor the material. The shape is exactly what the weighted catenary equation produces at those dimensions.
Steep at the base, flatter toward the top, the legs meeting at a point calculated before the first piece of steel left the Pittsburgh plant. You can run the equation yourself. Saarinen’s Arch is one of the few great public landmarks in America that is mathematically inevitable. Given the curve and the dimensions, there is no other shape it could be.
The next story is about how the federal government decided to build 41,000 miles of interstate highways across a country already connected, and what that cost the cities those roads ran through.
