The storm had arrived ahead of its forecast. At 12:23 on the morning of April 4, 1933, the USS Akron was flying at 1,600 feet over the Atlantic, her radio direction finders jammed by electrical disturbances, her crew unable to fix their position. Then the tail struck the water. The Navy’s investigation would later conclude that a downdraft of exceptional force had pushed the stern down before the control surfaces could compensate.

The ship was too low. The controls were not fast enough. The tail hit the ocean at full speed. What followed took about two minutes.

The impact tore through the lower fin, and gas leaked from the aft cells. The nose rose. The airship broke apart in the air and fell into the sea in pieces. There were no lifeboats aboard.

The Akron carried a limited number of life jackets, but in the dark, in water at 51 degrees Fahrenheit, with no rafts and no ships nearby, survival required luck and something to hold onto. The German steamer Phoebus was the nearest vessel. It reached the site minutes after the crash and pulled three men from the water. One of them was Herbert Wiley, the executive officer.

Rear Admiral William A. Moffett, who had boarded the Akron at Lakehurst the previous evening for a routine patrol along the Atlantic coast, was not among the survivors. Of the 76 men aboard, 73 drowned. It was the deadliest airship disaster in American history, and it remained so.

The question of how any of this came to be begins with a building. In 1929, a rubber company in Akron, Ohio, began constructing a structure so large that the weather inside it did not match the weather outside. The old photographs show it clearly: a steel arch rising 211 feet above the ground, as tall as a 20-story building, running 1,175 feet end to end across the Ohio floodplain. No columns inside.

No interior walls. Just one continuous steel envelope large enough to house the two largest flying machines America had ever built. On damp mornings, clouds formed near the roof. Workers on the upper catwalks watched rain falling inside the building while the sky outside stayed clear.

It had its own weather. And in its own way, it worked. For four years, the Goodyear Airdock sheltered the Navy’s most ambitious experiment in air power. Two airships were built inside it, one after the other, each longer than two football fields, each capable of launching and recovering fighter aircraft in flight.

Both were lost at sea. Of the 159 men aboard them, 75 died. The officer who survived the first crash commanded the second ship and survived that one too. Two men run through this story.

The first is Karl Arnstein, a Czech engineer who spent the First World War designing Zeppelins for Germany, came to Akron in 1924, and ended up designing both airships and the building in which they would be assembled. The second is Herbert Wiley, a Navy officer from Kentucky who served as executive officer of the first ship, was pulled from the Atlantic when she sank, took command of the second, and walked out of the Pacific alive when that one broke apart as well. The idea the Airdock was built to serve was among the strangest the American military ever funded on a large scale. In the years after the First World War, the Navy looked at the Atlantic and Pacific Oceans it was responsible for patrolling and faced a simple problem.

The oceans were vast. The aircraft it possessed had short range. A fighter operating from a carrier deck in 1925 had a combat radius of about 200 miles. A fleet carrier could steam at 30 knots.

Together, a carrier group could patrol a sector of ocean, but patrols meant movement, and the sector your carrier left behind was a sector no one was watching. What the Navy wanted was something that could stay in one position, loiter over a convoy for 40 hours without returning to base, and remain above a suspected submarine until it was forced to surface. If that platform could also carry its own fighter aircraft, it could extend its vision 400 miles in every direction without needing a friendly airfield or carrier within 2,000 miles. The platform in mind was a rigid airship.

Not the familiar blimp, but the type of structure the Germans had used to bomb London in 1915 and 1916: an aluminum framework covered in fabric, divided internally into gas cells, rigid enough to hold its shape in flight without relying on gas pressure for its integrity. German wartime Zeppelins had been about 650 feet long. The Navy wanted something bigger. Something that could carry a hangar for five biplane fighters inside its lower hull, with a mechanism to lower them through an opening in the airship’s belly and recover them in flight.

In 1926, the Bureau of Aeronautics issued a formal requirement. Two airships, each about 800 feet long, with a range of 10,000 miles and a minimum speed of 70 knots, carrying an internal hangar for five aircraft with a trapeze to launch and recover them while the ship was flying. Nobody had ever built anything like it. But the engineer who knew more about the internal structure of large rigid airships than anyone else in America had just arrived in Ohio.

Karl Arnstein was born in Prague in 1887. He studied structural engineering in Dresden, graduated in 1910, and went to work for the Zeppelin company in Friedrichshafen in 1915, the third year of the war. By then the German army was ordering Zeppelins as fast as they could be built, and the company’s engineering department was under constant pressure to produce ships longer, faster, and higher-flying than their predecessors. What Arnstein learned in those years was not the general principle of how to design an airship.

That principle was simple enough: a network of aluminum rings and longitudinal girders, covered in fabric, with gas cells inside. What he learned was the precise mathematics of keeping that lattice structure as light as possible while carrying the loads it had to bear. A Zeppelin in service was constantly being asked to fly higher and carry heavier loads, which meant constantly redesigning the same structural problem. How do you increase load capacity without increasing weight, while trying to reduce the weight of the structure itself?

The answer was geometry. An aluminum ring bent into an arch resists compressive stress far better than the same ring lying flat resists bending stress. An arch is a machine for converting weight into compressive force. Compression is something aluminum handles well.

Bending is something aluminum tolerates until it fails, and then it cracks. Arnstein spent nine years redesigning the load paths inside Zeppelin structures to carry more weight with less material. By the time Goodyear found him, he had performed those calculations on more rigid airships than any other living engineer outside Germany. In 1924, Goodyear’s managing director, P.

W. Litchfield, came to Friedrichshafen looking for German engineering talent to bring back to Akron. The Treaty of Versailles had forbidden the Zeppelin company from building military airships; the firm survived on design work and civilian contracts. Litchfield offered Arnstein the position of chief airship engineer at a salary no German company could match in 1924.

Arnstein accepted and moved to Akron with twelve other Zeppelin engineers. Over the next five years he designed small airships, worked on a small commercial airship, and waited for the contract that would let him build something on the scale of what he had learned in Germany. The Navy contract came in October 1928. Goodyear Zeppelin, a partnership Goodyear had formed specifically for the bid, won the contract to build both airships.

The first was designated ZRS-4, and would later be named USS Akron. The second was ZRS-5, later named USS Macon. The contract was worth about $5. 4 million for the two ships.

There was one problem the contract did not address, and it was the first problem Arnstein had to solve before anything else could begin. Where do you build an airship 785 feet long? An airship that size cannot be assembled outdoors. The gas cells had to be inflated in sequence as the structure rose, and gas cells in the wind behave unpredictably.

The outer fabric cover had to be fitted without moisture getting under it. The hydraulic systems, electrical wiring, and fuel lines all had to be installed and tested before the airship’s first flight, and none of them could tolerate rain. So the ship had to be built indoors, which meant a building at least 1,200 feet long, 325 feet wide, and 220 feet high, with no interior columns the airship could collide with. One vast opening spanning 325 feet at full height, with nothing in the middle to support it.

At that span, a conventional roof truss would weigh so much that the bearing walls required to hold it would have to be massive enough to consume the floor space the airships needed to maneuver. Arnstein looked at the problem and applied the same logic he had spent fifteen years applying to airship structures. Instead of a flat truss that resisted gravity with weight, he designed a parabolic arch that directed every load downward along its curve to the foundations. The arch bends from the floor on one side up to the summit and back down to the floor on the other side in one continuous steel curve.

At the summit it reaches 211 feet. At the base it meets a concrete footing that spreads the load through the Ohio clay beneath the floor. There is no tension in it. Every pound of the roof carries the steel in the same direction, pushing down and outward toward the foundations, and the arch redirects that thrust into the ground.

Think of a stone arch bridge. The bridge does not resist the weight of the traffic crossing it. It converts that weight into a downward thrust that travels along the curve to the abutments at either end. The arch stands because every stone in it is in compression, and stone handles compression better than almost anything.

Arnstein built the same machine out of steel, on the scale of a small mountain, over a floor that had to remain clear of supports. The arch he designed used 9,000 tons of structural steel, prefabricated in Pittsburgh mills and shipped to Akron by rail. On site, temporary wooden scaffolding held each piece while it was bolted into place. The arch grew from both sides at once, the two halves curving upward toward each other like hands meeting across the building.

Every section was bolted to the section below it and supported by scaffolding until the final piece was set at the summit. When that last piece was bolted, the arch became self-supporting for the first time, and the scaffolding was removed. That closing moment turned every stressed element in the arch from tension to compression. Every joint that had been trying to separate was now pressed together.

It was the moment a partial construction became a complete structure, just as the keystone in a Roman arch turns two leaning columns into a bridge. Arnstein had made the same calculations hundreds of times designing Zeppelin structures. Here he was applying them at the scale of a building, in the spring of 1930 in Ohio, where 9,000 tons of steel did exactly what the geometry predicted. Construction ran from late 1929 into the spring of 1930.

The completed building was 1,175 feet long, 325 feet wide, and 211 feet high at the peak of the arch. It enclosed about 55 million cubic feet of air. Its total cost came to roughly $3 million. Then the weather asserted itself.

A building that long contains enough air to generate its own local climate. In an Akron summer, that air is warm and humid. The steel envelope conducts the sun’s heat on its southern face while the northern face stays cooler. At night, the whole envelope radiates the day’s heat, and the interior temperature drops faster than the outside air.

On damp mornings, that differential is enough to push the air near the roof past its dew point. Two or three hundred feet up, the temperature may be two or three degrees below the temperature at the floor. That is quite enough. Moisture condenses into clouds.

Workers on the upper platforms reported walking through mist. Rain fell inside the building. Small puddles formed on the concrete floor and dried by mid-morning as the building warmed through the day. Arnstein recorded this in his reports in the same practical tone he used for load calculations.

A building this size contained enough atmosphere to create its own weather system. He managed it by scheduling moisture-sensitive work, such as the installation of gas cells, electrical wiring, and fabric covers, for the warmest part of the day when the fog cleared. There was no way to make a 55-million-cubic-foot building behave like an ordinary room. The sky does not stop at the walls.

It just has a roof. Construction of the USS Akron began inside the building while the hangar roof was still being erected around it. The hull rose ring by ring, each aluminum ring assembled on the floor and lifted into place by a crane. Longitudinal girders running bow to stern were fitted across each ring as it was raised and connected to the one before.

By early 1931, the twelve gas cells were installed, the outer fabric cover was in place, and the airship, which had never touched the sky, was longer than any structure inside any building in the country. On August 8, 1931, a ground crew of hundreds walked the USS Akron out of the Airdock, and she flew for the first time. The man watching from the ground as she rose was Karl Arnstein. The man who came aboard that autumn as executive officer was Herbert Wiley.

Wiley had been in the Navy since 1910, had moved into lighter-than-air aviation in the early 1920s, and by the time the Akron was commissioned on October 27, 1931, had more flight time in rigid airships than any other officer. He was 44 years old, from Maysville, Kentucky, and precise in the manner of men who have spent years handling the physics of large, slow aircraft. An airship is not a machine that forgives inattention. It requires constant monitoring: gas pressure, ballast levels, engine temperatures, weather reports from stations ahead, all of it simultaneously understood by one officer.

Wiley understood all of it. The ship he served on was remarkable by any measure. The Akron was 785 feet long, longer than the original Empire State Building was tall. Her twelve engines produced a total of 4,480 horsepower and drove her at 84 miles per hour.

Her gas cells held 6. 85 million cubic feet of helium, enough lifting force to carry 403,000 pounds. She carried a crew of more than 60, with capacity for additional passengers, fuel for 10,000 miles, and in the cavity of her lower keel, an iron trapeze that made her something new. The trapeze mechanism was the part of the design that made the Akron a flying aircraft carrier rather than merely a huge patrol ship.

The trapeze lowered from the internal hangar through a T-shaped opening in the hull. A pilot flew a Curtiss F9C Sparrowhawk toward a target moving in three dimensions at once. He had to match the speed of something that swung and turned while he himself swung and turned, trying to hook a latch onto a bar located somewhere near where he expected it to be. The pilots who mastered it described it as one of the hardest things they ever learned to do in an airplane.

For the Navy watching in 1932, it was a demonstration of something never done before. No navy in history had launched and recovered aircraft from an airborne platform. The Akron was doing it routinely. The man who championed the program that made this possible was Rear Admiral William A.

Moffett, chief of the Navy’s Bureau of Aeronautics since its founding in 1921. Moffett had fought battleship admirals for a decade to keep naval aviation funded, and had staked his career on both the carrier program and the rigid airship program at once. He was 68 in the spring of 1933. On April 3, he boarded the USS Akron at Lakehurst, New Jersey for a routine patrol along the Atlantic coast.

Weather reports that day indicated a front approaching from the south. The forecast called for it to reach the New Jersey coast the following morning. The airship departed early in the evening, heading south along the coast. By midnight the storm was moving faster than expected.

The ship was at 1,600 feet over the Atlantic in severe turbulence. Atmospheric electricity jammed the radio direction finders. The crew could not get an accurate fix on their position. At 12:23 on the morning of April 4, 1933, the tail struck the water.

The Navy’s investigation attributed the disaster to the unexpected severity of the storm and the lack of weather information available to the crew. The report noted that the Akron had carried no life rafts and recommended that all future ships be equipped with them. The report did not recommend ending the program. It recommended better weather support and proceeding with the second ship.

In the Airdock at Akron, the second ship was nearly complete. The USS Macon had been under construction since 1931, in parallel with the Akron’s operations and first flights. She was commissioned on June 23, 1933, eighty days after the Akron went down in the Atlantic. The Navy did not slow the program.

It accelerated it. The officer assigned to command the Macon was Herbert Wiley. He took command in October 1933 and flew the ship to Moffett Field in California, the base the Navy had recently named for the admiral who had gone down with his predecessor’s ship. Consider the weight of that assignment.

The Navy had lost 73 dead and one destroyed ship. It commissioned a replacement and gave command to the only senior officer who had been on the Akron’s bridge and come out of the ocean alive. Wiley knew more about what a ship of this design could do, and could not do, than almost anyone in the Navy. The reasoning was that if anyone could command the Macon and spare her the fate that took the Akron, it was Wiley.

From Moffett Field, the Macon flew patrol missions over the Pacific and continued developing the Sparrowhawk operations the Akron had pioneered. By the end of 1934, Wiley’s crew could launch and recover all five aircraft in one continuous sequence. Each pilot flew toward the trapeze, hooked his plane, and was pulled into the hangar while the next aircraft was lowered before the previous one’s engine had stopped. The reports from those flights read like a program about to fulfill every promise it had made.

Then a fin failed. The Macon had suffered minor fin damage in an airborne incident in April 1934. The upper fin was inspected and repaired at Moffett Field. The inspection report described the repair as satisfactory.

The word satisfactory meant the immediate damage had been addressed. It did not mean the surrounding structure had been examined for stress. It did not mean the repaired section matched the original strength. It meant the visible damage was fixed and the inspector had approved it.

On February 12, 1935, the Macon was flying over the California coast at about 1,500 feet, returning to Moffett Field after fleet exercises off Monterey. The weather was cloudy but not severe. At 5:07 in the afternoon, a gust of wind struck the upper fin. The repaired section failed first.

The failure spread inward across the structure adjacent to the repair, through material that showed no visible signs of damage and had not been replaced. The fin tore away from the hull, and its separation punctured the gas cells in the stern. The stern began to lose lift. Gas leaked through the internal spaces of the structure and reached the cells toward the bow.

The ship was out of control. She climbed to 4,850 feet as the crew dumped fuel and ballast in an emergency. Then she fell. Wiley ordered the life jackets distributed and the crew to their emergency stations.

The water off Point Sur was 56 degrees Fahrenheit and clear. The Macon struck the surface at 5:15 in the afternoon. Of the 83 men aboard, 81 survived. Two men died, one from injuries, one drowning in the first minutes before surface ships arrived.

A Coast Guard cutter, the Macon, which had been conducting exercises nearby, reached the survivors within 20 minutes. Herbert Wiley came out of the Pacific alive for the second time. The congressional hearings that followed were longer and less forgiving than the Akron investigation. The Navy’s airship advocates argued the same case.

Both losses were caused by weather and inadequate maintenance, not by inherent design flaws. Better inspection standards would have prevented future losses. The program had proved its intrinsic capability. Losing two ships to storms and metal fatigue was no reason to abandon a proven technology.

The hearings lasted weeks. At the end, the rigid airship program received no further appropriations. The two ships Arnstein had designed, and the building he had designed to build them, had been made for a program that no longer existed. Arnstein submitted written testimony arguing that the structural failure in the Macon was a maintenance failure, not a design failure.

The accident investigation board agreed with his analysis of how the fin had failed. That agreement did not change the outcome. Congress had made its decision before the hearings began, and the hearings were paperwork for a decision already made. He returned to Goodyear.

The Airdock was still there, the largest open-floor building in the country, still making its own weather on damp mornings over a floor that no longer held anything to assemble. Goodyear began using the building to build non-rigid airships almost before the program was formally ended. When the Second World War broke out, the Navy wanted non-rigid airships for anti-submarine patrols along the Atlantic convoy routes. The building designed to house two flying aircraft carriers spent the war building 135 airships to escort convoys.

At the height of the war, the Airdock employed thousands of workers on round-the-clock shifts. The arch Arnstein had closed at the summit in 1930 looked down on all of it in silence. Arnstein stayed at Goodyear until 1957. He became the company’s chief structural engineer, overseeing not just airships but the full range of structural work the company required.

He designed rubber-processing equipment and industrial buildings. He did not speak publicly, according to published records, about the loss of the two ships. He retired in 1957 and died in Akron in 1974 at the age of 86, outliving the Akron and the Macon by nearly four decades. His name appears in the engineering histories of the multi-arch dam, the Sparrowhawk trapeze, and the parabolic arch over the Airdock floor.

The clouds that formed at 200 feet do not appear in the engineering records. They appear in the accounts the workers left behind. Wiley returned to surface ships. He commanded ships through the Second World War, rose to rear admiral, and retired in 1947.

He died in San Diego in 1960. The two accidents are the first line in every biography of him, as if a man who walked out of two burning buildings would be defined by the fires rather than by everything he did afterward. The Airdock passed through several owners: Goodyear, then Loral, then Lockheed Martin, then the state of Ohio for a period, then back to Lockheed Martin for classified government work. The National Register of Historic Places listed it in 1994.

In 2012, a company developing hybrid cargo airships moved into the building to assemble its prototypes, and for a few years the Airdock housed an airship program again, for the first time since the war. That project ended without producing a commercial aircraft. The building was still standing when the project ended. The parabolic arch still holds.

It has had no major structural modification in 95 years, apart from routine maintenance. The 9,000 tons of steel Arnstein put under compression in 1929 and 1930 are still under compression. That is the whole logic of the arch. If you load it correctly, it stands indefinitely.

You can see it from Goodyear Boulevard in Akron. The building is not open for public tours, but it is visible from the road. What you see from the road is the building itself. Walk along the fence line and watch how the scale changes when a car-sized object appears at the base of the arch.

The structure that looked like a massive industrial building from a distance becomes something else. A building that covers more floor space than most city neighborhoods. An arch spanning 325 feet in a single opening with nothing in the middle to support it. The last Sparrowhawks that flew from the Akron and the Macon are in museums now.

One hangs from the ceiling at the National Naval Aviation Museum in Pensacola, Florida. It is a small biplane with a hook above its fuselage. That hook is the piece the entire program depended on: the point of contact between a fighter flying under an Atlantic sky and a 785-foot ship flying above it. The hook connected a Sparrowhawk to the trapeze bar, and the bar pulled it through the opening into the belly of the airship.

For a few years in the early 1930s, that process happened routinely over the Pacific coast, proving that the mission for which the airships were designed was in fact possible. Both ships that proved it now lie at the bottom of the ocean. The building they were made in still stands. The Airdock cost $3 million in 1929.

The two ships cost another $5. 4 million. The program built for all of this lasted four years, lost every ship it operated, and killed 75 men. No one who designed it, funded it, built it, or commanded it ever profited from it.

What they left behind is the largest parabolic arch built in the United States up to that time. An arch that still carries its own weight, and the weight of everything above it, on the flat Ohio ground where Karl Arnstein closed its summit in the spring of 1930 and watched the scaffolding come down.