Building Apollo

A review of Angle of Attack: Harrison Storms and the Race to the Moon

(Unless otherwise noted, all quotes are from Angle of Attack)

The Apollo program looms large in the American psyche. Nearly every American knows who Neil Armstrong is, and he was voted one of the 100 most significant Americans of all time by Smithsonian Magazine. The photos taken during the Apollo missions remain some of the most famous in the world. In a 1999 Pew survey that asked about the greatest American achievements of the 20th century, more people said the space program than any other answer. More than 50 years later, Americans continue to feel pride that no other country has managed to put a man on the moon.

Enthusiasm for the Apollo program tends to be focused on the astronauts who piloted the spacecraft, and on the NASA mission control staff that managed the flights from the ground. Comparatively less focus is placed on the actual construction of the Apollo spacecraft and the rockets that put them into orbit. Everyone knows who Neil Armstrong is, but almost no one knows who built the Eagle lander that carried him to the lunar surface (it was Grumman Aerospace). In fictional treatments like the movie First Man, the rocket is simply there, ready and waiting for the astronauts to take their historic flight.

But the astronauts, and NASA, were just the tip of an enormous iceberg of industrial infrastructure, made up of 400,000 workers and 20,000 individual contractors that designed and built the various rockets and spacecraft of the Apollo program.

Angle of Attack: Harrison Storms and the Race to the Moon, by Mike Gray, is a book about one of these contractors, North American Aviation, and the man, Harrison Storms, in charge of the company’s Apollo efforts. It tells the story of what it took to get the rocket from design sketches to the launchpad, the blood and sweat required to build a spacecraft capable of traveling hundreds of thousands of miles through the vacuum of space, landing astronauts on the moon, and returning them safely to earth.

Rise of North American Aviation

North American Aviation was founded in 1928, but for all practical purposes the company's real beginning was several years later, when General Motors bought a controlling share, merged it with the Fokker aircraft company, and put James “Dutch” Kindleberger in charge of the operation. Under Kindleberger, North American built a series of wildly successful airplanes. In 1935, it introduced the T-6 Texan, “one of the most successful training planes in history," with more than 15,000 built in various configurations. In 1940, it brought the P-51 Mustang, considered “the finest combat fighter of its generation," from first sketches to a flying prototype in just 100 days. The next year, it introduced the B-25, the bomber used in the “Doolittle Raid” against Tokyo in response to the Japanese attack on Pearl Harbor, and the most produced American medium bomber of the war. By the end of World War II, North American had built more than 42,000 aircraft.

Harrison Storms joined North American in 1941, six months before Pearl Harbor. Storms grew up building model airplanes, and after getting a masters in Mechanical Engineering from Northwestern, he went to Caltech for a second masters in aeronautical engineering, studying under the famous Theodore von Karman. While working in Caltech’s wind tunnel, Storms was offered a job at North American. Over the next 15 years, Storms played a major role in the design of half a dozen aircraft, including the F-86 Sabre (which remains the most produced US jet fighter) and the F-100 Super Sabre (the first US fighter capable of level supersonic flight). In 1957, he was promoted to Chief Engineer of North American’s LA division. 

As the Soviet threat loomed in the 1950s, the US developed a series of increasingly advanced aircraft to push the boundaries of aerospace technology and maintain its military edge. At North American, Storms led the development of two of these projects, the B-70 and the X-15.

The B-70 was a high-speed bomber designed, according to Ed Rees in The Manned Missile, to give the US “unchallenged military supremacy." It would cruise at an elevation of 70,000 feet at Mach 3, outrunning any possible interceptor aircraft as it flew deep behind Soviet lines to deliver its payload. The critical breakthrough in the B-70’s design was the utilization of “compression lift.” With the proper shape, a hypersonic aircraft would ride on its own supersonic shockwave, increasing lift by almost 30%. But this was far from the only advance: achieving the performance the B-70 was designed for required breakthroughs in “a dozen dark arts.” (Rees p10) A new type of brazing process, for instance, had to be invented to create the lightweight steel honeycomb panels that made up the body of the aircraft, and a special, heated contoured form was invented to bend the panels into the proper shape (Rees 23). To minimize weight, the steel needed to be rolled as thin as possible: two thousands of an inch in some cases, (Rees p24) which required inventing new ways to roll and shape steel. The steel fabrication techniques created for the B-70 were, according to Rees, “as great an advance in the state of this art as the plane itself represents in its design." 

The B-70 program would ultimately be canceled amidst great debate in the 1960s. As antiaircraft missiles continued to improve, the idea that a fast enough bomber would be unstoppable appeared increasingly unsound. But during the program’s lifespan, North American built two prototypes, pushing the aerospace technology frontier forward as it did so. Designed and built as the US space program was just coming online, the advances of the B-70 were among those that would make later manned spaceflight missions possible.¹

B-70 prototype

The other advanced aircraft project led by Storms was the X-15, a research aircraft for NACA (NASA’s predecessor) for investigating hypersonic flight. The X-15 was, in some ways, a piloted missile: it mounted a rocket engine to a long rod of Inconel-X, with two stubby little wings mounted to the side. The X-15 flew so high that control surfaces wouldn’t work in the thin air, and small rocket motors were required to steer it. And it didn’t take off or land anything like a normal plane: it “took off” by being dropped from the wing of a B-52, and landed using two “skids” rather than rear wheels. The X-15 would ultimately achieve a speed of Mach 6.7 (4,520 miles per hour), a record which remains unbroken for crewed, powered aircraft.

Like the B-70, building the X-15 required pushing the state of aerospace technology forward, including developing new fabrication techniques. Inconel X, for instance, was a new alloy, and methods had to be invented for working it. To minimize weight, the metal had to be machined to tolerances of 0.001 inches, which X-15 test pilot Scott Crossfield described as “like making a Stradivarius, if not more delicate." New welding technology likewise had to be developed to minimize stresses and achieve the required levels of precision:

“When you weld two pieces of metal together, each is subject to varying temperatures from the welding torch. As the torch moves along, the new area heats up while the one just passed cools. Thus there are stresses and strains in the molecular structure of the metal undetectable to the naked eye. By placing the entire structure in an oven after welding and raising the temperature to 1900 degrees we were able to cool it uniformly, ironing out the strains…After a fine polishing, the hundreds of welds were impossible to locate with the human eye. The wing looked like one solid piece of smooth metal.”

Scott Crossfield, Always Another Dawn

North American X-15

Working out these welding techniques and other fabrication methods took years, but eventually North American learned how to weld, machine, and shape Inconel X into an aircraft capable of traveling thousands of miles an hour. Crossfield argued that knowing how to work with Inconel X alone was worth the entire cost of the X-15 program. Writing in 1960, Cross predicted that “all future space projects will benefit directly or indirectly from our work with Inconel X." (This speculation proved correct, and Inconel X would indeed be used in several components on the Apollo program.)

North American and the rocket game

While Storms was pushing the boundaries of aircraft performance with the B-70 and the X-15, the same process was happening elsewhere in North American, with rocket engines.

North American had been working on rocket engines since the 1940s. During WWII, Germany was the unquestioned leader in rocketry, and by 1944 had developed the V-2, a liquid-fueled guided missile that could send a 2000-pound payload 200 miles.² Germany used the V-2 to rain down seemingly unstoppable destruction on London:

“The V-2 had a maximum speed greater than 3800 mph (nearly ten times the maximum speed of a P-51 Mustang). No fighter could catch it and no anti-aircraft artillery could target it. On radar sets of the period it was a line not a dot. And since it was outrunning its own sound, the V-2 hit without warning.”

The Navaho Missile Project

In 1945, North American won a contract to develop a missile superior to the V-2. Because the Germans were miles ahead, North American started by learning everything it could about the German rocket program, hiring engineers who were familiar with the V-2, and studying captured research and equipment.

Though North American was contracted to build a superior V-2 with a 500-mile range, over the next several years the project evolved into an intercontinental missile capable of traveling 5000 miles. Using the V-2 engine as a starting point, by 1956 North American had built a two-stage cruise missile designed to hit targets in the Soviet Union from bases in the US. The Navaho missile was first boosted by a V-2 derived rocket engine to its cruising speed and altitude, after which it traveled the rest of the way by ramjet.

The final Navaho missile launch in November 1958

The Navaho itself ended up being “outrun by its own technology." By the time it was making test flights, the rocket technology it pioneered had made its way into the Atlas, a missile which would be fired into space and drop its warhead on its target in a ballistic arc, far superior to a cruise missile, which had to travel thousands of miles through the atmosphere. North American also struggled to get the Navaho to work properly (the first 11 launch attempts failed), lending it the nickname “Never-Go-Navaho."

But like with the X-15 and the B-70, simply building the Navaho resulted in major technological advances. In his book about the project, James Gibson describes the Navaho as “the least known, but most important of the United States early missile programs," producing “major advances in every discipline in engineering.”

“The technology Navaho created is worth more than all the gold in Fort Knox. In rocket technology alone the Navaho made possible the Thor, Jupiter and Redstone missiles. It also allowed the construction of the Atlas ICBM’s engines, thus sealing its own fate. In the years to come this engine technology would power the Apollo moon rocket, and our present Space Shuttle. Electronics was another area enhanced by Navaho technology. The program gave us the airborne digital computer, modular electronic circuitry, and the all inertial navigation system. The development of modular [transistorized] circuitry alone revolutionized the electronics industry, improving the reliability and reparability of countless electronic devices. The Navaho even affected manufacturing. Probably the greatest legacy is Chem-Milling, now used throughout our commercial aerospace industry for the manufacturing of aircraft body panels. Automatic Tungsten arc welding and lightweight bonded aluminum honeycomb were also Navaho developments.”

The Navaho Missile Project

In 1955, North American consolidated its rocket development efforts into a single division, Rocketdyne. Rocketdyne engines, based on its Navaho missile work, would go on to power hundreds of Atlas and Thor ballistic missiles, as well as the Redstone rocket that would launch the Mercury spacecraft. Every major engine used on the Saturn V, including the monstrous F-1 engine that powered the first stage, was a Rocketdyne engine.

The space race

As Storms was building the X-15 and B-70 and Rocketdyne was building the engines for America’s ICBMs, the space race was heating up.

America’s space program can be traced to Germany’s V-2 program. The program was led by Wernher von Braun, who had a team of several thousand engineers and scientists researching and building rockets at Peenemünde.³ When it became clear that Germany was on the brink of losing the war, the question for von Braun’s team became who to surrender to, and where they could best continue their rocketry work. The Americans seemed preferable to the Soviets, but Peenemünde was very close to Soviet lines, forcing von Braun to move his entire organization hundreds of miles across war-torn Germany to reach the Americans. Ultimately, von Braun and around one hundred other key personnel were brought to the US as part of Operation Paperclip and set up in Huntsville, Alabama, where they began to build missiles for the US Army.

(Another, smaller group of German rocket scientists would end up in Soviet hands. When the Soviets announced their first intercontinental ballistic missile in 1957, on the heels of American ICBM launch failures, one US government official lamented that “we captured the wrong Germans.")

But von Braun wasn’t passionate about rockets for their own sake. He advocated for space travel in a series of books and articles in the 1950s, including one about how to reach Mars, The Mars Project. But Eisenhower, worried about setting a precedent for militarization of space, had decided that the Navy would launch a non-military satellite using the Vanguard rocket, beating out the Army to be the first US satellite into orbit.

This didn’t sit well with von Braun, who knew that Soviet rocketry was rapidly advancing, and that the Vanguard was beset with problems:

“Von Braun rolled up the blueprint and handed it to O’Keefe. ‘We are now watching through radar the firings of Russian ballistic missiles,’ he said. ‘We know they have the capacity to go into space. Now, I will tell you what is going to happen. During the spring of next year, the Vanguard rocket will get into trouble. It will get into worse and worse trouble and finally its firing will be delayed. In the meantime, the Russians will fire and they will get their thing up before we do.’ O’Keefe was shaken. He had no trouble grasping the political significance of being beaten into space by the Russians. ‘What can I do?’ ‘I want you to go see John Hagen at the Naval Research Laboratory, and I want you to tell him that if he wants to, he can paint “Vanguard” right up the side of my rocket. He can do anything he wants to, but he is to use my rocket, not his, because my rocket will work and his won’t.’”

Von Braun was unable to convince the powers that be. At one point, government officials were so worried that von Braun might “accidentally” launch ahead of the Navy that they inspected his missiles to be sure he didn’t add a fourth stage to them. But von Braun unfortunately proved prescient, as the Soviets launched Sputnik in October 1957, ahead of the Americans. Von Braun begged to be allowed to launch. “Vanguard will never make it! We have the hardware on the shelf for God’s sake! Turn us loose and let us do something!” He was once again ignored. Two months later, the Vanguard rocket exploded on the launchpad (in an incident dubbed “Kaputnik” by the press), and the US finally let von Braun off the leash. In less than 90 days, von Braun and his team reworked their Jupiter-C missile to make it capable of reaching orbit, and successfully launched the first US satellite, Explorer 1, in February 1958. The space race was on.

The Soviets quickly racked up more successes (launching a live dog, Laika, into space on Sputnik II 30 days after the first Sputnik), and America scrambled to marshal a response. By April, there were 29 different congressional bills and resolutions addressing the space effort, and every branch of the armed forces was making its own plans for spaceflight. Harrison Storms, who had long been interested in space, tried to jump in right away. His idea was simple: take the X-15, which was already basically a spacecraft, and strap it to spare rocket boosters from the Navaho missile project, which had recently been canceled. Storms’ idea, which he had already been toying with prior to Sputnik, became known as the X-15B, but North American was unable to convince the Air Force of its merits. The Air Force thought that a manned capsule was a better bet than some kind of space plane, and solicited proposals for one as part of their crash Man In Space Soonest (MISS) project. The North American team, led by Storms, won the contract, but MISS was canceled when Eisenhower declared that space exploration would be a civilian matter, not a military one.

Instead, a new organization was created, NASA. The core of this new entity would be former aeronautics research organization NACA, which would merge with rocketry efforts taking place in different parts of the armed services. The new NASA received the Army’s jet propulsion laboratory in Pasadena. It got the enormous F-1 rocket that Rocketdyne had been building for the Air Force. And it got von Braun’s group in Huntsville, which became the Marshall Space Flight Center. MISS was replaced by the Mercury Program, with the goal of putting a US astronaut into space.

North American, Storms, and the pivot to space

North American bid on the proposal to build the Mercury spacecraft, but the contract ultimately went to McDonnell Aircraft. Despite Storms’ interest in space, the rest of North American was less interested. Kindleberger didn’t like the space business (he had been against the X-15 project as well, and had only pursued it as Storms’ urging), and neither did president of the company Lee Atwood. As a mere engineering vice president, there wasn’t much that Storms could do.

But in 1960, Storms got his chance to put North American into the space business. Kindleberger asked him to move to North American’s missile division to try and generate some new business: at the time the division had just a single contract for the Hound Dog missile. Instead, Storms convinced Kindleberger to let him run the entire division.

When Storms arrived, he found the division in shambles: the facilities were a wreck, and the Hound Dog program was struggling. He wasted no time. First, he got the Hound Dog program back on track. When Storms first arrived, the missile “was aerodynamically unstable, and couldn’t hit the target." Storms suggested adding weight to the nose to make it less tail-heavy (a trick he picked up from his model-building days). The next test, the missile “flew like an arrow to the target."

Next, Storms began to assemble his team, raiding the rest of North American for the best talent he could find. “His concept was simplicity itself: surround yourself with brilliant people from every conceivable discipline, get them all facing the same direction, then build a fire under their ass.” He cajoled people with the possibility of going to the moon, and the opportunity to finally work on something other than weapons designed to kill people. Storms didn’t take no for an answer: when an engineer said he couldn’t join because he needed 9 months to finish his PhD thesis, Storms assigned an entire engineering team to assist on the calculations, and the thesis was finished the following week. To give the division a more ‘scientific aura,’ he created an “Advisory Group” of some of the nation’s leading scientists, enticing them with consulting fees and several trips to Southern California a year.

For equipment, Storms stole “everything that isn’t nailed down” from his former LA division, taking entire laboratories down to missile division headquarters. He spent millions of dollars, far beyond his allotted budget, to renovate the facilities to create the proper image, renaming the division “Space and Information Systems." to appeal to the computer and electronic experts he needed to recruit.

While Storms was turning North American’s missile division into a spacecraft manufacturer, the space race lurched forward. When Kennedy was elected in 1960, his appointee to lead NASA, Jim Webb, vowed to put the US ahead. But the Soviets continued to rack up victories. In 1959 they (accidentally) put the first spacecraft into orbit around the sun, and in 1961 they successfully launched a probe to Venus and put the first man, Yuri Gagarin, into orbit around the earth.

In response, Kennedy aimed for a target the US could hit before the Soviets. In a statement before a joint session of Congress, he announced the audacious goal of putting a man on the moon by the end of the decade:

“Recognizing the head start obtained by the Soviets with their large rocket engines, and recognizing the likelihood that they will exploit this lead for some time to come in still more impressive successes, we nevertheless are required to make new efforts on our own…I believe this nation should commit itself to achieving the goal, before this decade is out, of landing a man on the moon and returning him safely to earth.”

The Apollo Program

Even before Kennedy’s announcement, NASA had been busy with plans for a moonshot. In 1960, NASA solicited proposals for basic concepts for the design of a lunar spacecraft. (It then took the best ideas from these concepts, said “thanks but no thanks” to the contractors, and proceeded to use its own design, which became the design for Apollo.) At the same time, von Braun’s group was working out the design for the massive Saturn V rocket that would put the spacecraft into orbit. By the time of Kennedy’s speech, these designs were nearly to the point where they could be put out for bid by contractors.

The Saturn V went out to bid first, with each of its three stages bid separately. Storms zeroed in on the second stage, by far the most complicated of the three stages. While the first stage would use conventional kerosene as a fuel, the second and third stage would use the more difficult to handle (but more efficient) liquid hydrogen.⁴ And to meet the weight requirements, the second stage structure would have to be almost unbelievably light: just seven percent of the weight of the stage, including the engines, could be the rocket itself, with the other 93% going to fuel.

To achieve this, Storms’ team proposed an audacious design. The conventional way of building the rocket was to have the fuel (hydrogen) and the oxidizer (oxygen) in two separate tanks. But on the S-II, North American proposed a “common bulkhead,” using the bottom of one tank as the top of the other tank. By cutting out the rounded top of one of the tanks, the common bulkhead would dramatically reduce the weight of the stage, at the expense of massively increased fabrication complexity.

Common bulkheads had been built before (and would be used on the Saturn third stage), but none of this size and lightness. Many at North American, and at NASA, doubted that such a thing could be built:

“The production engineers — the people who would actually have to build this monster — went into cardiac arrest. Nobody had ever made a featherweight high-strength dome of such vast size, and insulating the thing would be a nightmare. At 300 degrees below zero, the liquid oxygen on one side of the thin shell would be 100 degrees hotter than the liquid hydrogen only an inch or two away…Even if they could solve the structural problems, how the hell would they build it? It couldn’t be solid metal; that would weigh too much. It would have to be made out of some kind of metallic honeycomb — cellular material sandwiched between two thin sheets of metal. Storms had experience with honeycomb structures on the B-70. But the strength of honeycomb depends on an absolutely perfect bond between the honeycomb sandwich and the metal sheets on either side; any imperfection and the whole thing turns to Jello. And in this case the honeycomb structure would not be some simple box like a wing or a tail, it would be this immense thirty-foot dome, as thin for its size as an eggshell, its surface a constantly changing compound curve.”

Storms listened to the objections, but saw no alternative. Their proposal for the S-II would proceed with a common bulkhead.

But the real prize of the Apollo program wasn’t the rocket, but the Apollo spacecraft itself. Storms decided to bid on that too, though the odds seemed long. Many doubted that NASA would be willing to load up a single contractor with two large portions of the program, and thanks to the Mercury program other companies, like McDonnell, had much more experience building space capsules. But Storms threw out all the stops on the proposal, spending millions of dollars in man-hours (and once again going massively over budget) putting it together. He’s only saved by the fact that the slow, punch card-based accounting of the time meant that it would take at least 60 days for corporate headquarters to realize how much he was spending, by which time the proposal would be in.⁵

To the surprise of many, North American won both proposals. By the time of the proposal, North American had a successful history working with NASA/NACA on the X-15, and many of the pilots within NASA were strongly predisposed toward North American due to the quality of their aircraft. During the scoring process for the spacecraft proposals, Alan Shepard stated, “this is all a waste of time…It doesn’t make any difference what the score is. North American is going to win.” Gus Grissom likewise said, “By God, I’m going to do everything I can do to make sure North American gets this contract.” In fact, North American’s spacecraft proposal was scored slightly lower than Martin’s, but North American was chosen anyway, as it was thought it would do the best job. This fact would later come back to haunt North American.

After years of trying, Storms had finally broken into the space business.

Building Apollo

After winning the proposal, North American was faced with the challenge of actually building what it said it would. Storms quickly staffed up the space division from 7,000 workers to nearly 30,000 in 18 months.

The second stage proved just as nightmarish as the engineers had predicted. There’s an old engineering saying: “anyone can design a bridge that stands up, but only an engineer can design a bridge that barely stands up." It takes engineering skill to trim away unnecessary structure and provide something that meets, but doesn’t exceed, the requirements.

Nowhere is this sort of trimming more important than in spaceflight. Every bit of mass added requires more fuel to lift it, which then requires even more fuel to lift that fuel, and so on, a vicious cycle known as the “tyranny of the rocket equation." On the Saturn V, every pound of payload added meant an additional 75 pounds of rocket and fuel to lift it into space. To get to the moon, engineers would need to cut out as much weight as they could from the spacecraft and the rocket.

North American ended up shouldering an enormous amount of this weight reduction burden. The lunar lander itself was originally planned to be 20,000 pounds, but it quickly ballooned in size as the design was fleshed out, and would ultimately come in closer to 33,000 pounds. The Apollo spacecraft itself got larger too. The only way to compensate for this was to cut weight from the rocket itself. 

Lunar lander weight over time, from “Moon Lander.”

There wasn’t much to be gained in cutting weight from the first stage: because it was at the bottom and got discarded relatively early in flight, any weight reductions there had less impact. Cutting a pound from the first stage only added about 0.07 pounds in payload capacity. And because of how the project was planned, the third stage was far along the design and construction process by the time North American started work on the second stage. (The Saturn V third stage was based on the second stage of the earlier Saturn I rocket, which was already under construction by the time North American was awarded the contract.)

That left the second stage as the place to trim weight as the lander and Apollo spacecraft got bigger. By the end of the program, the second stage of the Saturn V would be the most efficient structure ever built.

Getting the weight down required creativity. For the body of the rocket, Storms chose an aluminum alloy, 2014 T6, that got stronger as it got colder, taking advantage of the fact that the liquid hydrogen fuel would be at 400 degrees F below zero. This meant the tank walls could be 30 percent thinner.

But this also meant the insulation that would keep the fuel cold had to be on the outside of the tank, rather than the inside. The engineers had to find a way to create a perfect bond between the insulation and the super cold walls of the tank: if there were air pockets, the low temperatures would liquify the oxygen in the air, which would then break down the bond between the insulation and the tank, causing the insulation to peel away. After months of struggling to attach panels of insulation, North American eventually developed a way to spray insulation directly on the tank walls, preventing air pockets from occurring.

Once it had selected the material for the tank walls, North American had to figure out how to weld it. The welds on the second stage needed to be accurate to within 0.01 inches, and flawless: even a speck of dust could potentially cause a crack that would lead to catastrophic failure. “At a time in history when a flawless weld of a few feet was considered miraculous," Gray notes, “the S-II required half a mile of flawless welds.” The fact that 2014 T6 aluminum was extremely difficult to weld did not help.

To achieve this, North American ultimately had to do its welding in essentially a clean room. Welders would enter through a double airlock, wearing lint-free clothing. Floors were constantly mopped to remove dust, eating and smoking was forbidden, and the humidity level was carefully controlled. Because the level of precision required was human capacity, automatic welding machines were invented. And because the components were so large and thin, special assembly tools were built to hold them in place:

“On their first attempt to join two cylinders together, they were 80 percent of the way around the seam when the remaining section suddenly ballooned out of shape from the heat buildup. After that, each approach they tried got progressively more complex and muscular until finally the assembly tools were so immense they simply overwhelmed the problem with brute force. Each ring was enclosed in a massive precision jig that had 15,000 adjustment screws placed an inch apart around the whole hundred-foot circumference. The cylinders were then mounted on a giant turntable that moved the seam past stationary weld heads with micrometric accuracy.”

Unsurprisingly, achieving this perfect welding was unbelievably difficult, and as late as 1968 North American still had trouble performing some welds reliably.

Fabricating the common bulkhead, another weight-reducing design choice, also proved incredibly difficult. NASA, skeptical that it could be built, insisted that North American consider other options until very late in the program. The common bulkhead required precisely forming thin sheets of aluminum into complex curves on a scale that had never been done before:

“Each dome was made of a dozen gores — immense pie-shaped wedges of aluminum eight feet wide at the base and twenty feet from there to the apex. They had to be perfectly curved in two directions — a spherical curve from side to side, and a complex double ellipsoid from the base to the apex. No techniques existed for precision forming of such large unwieldy slabs; there was no hydraulic press in the country equal to the occasion. So they started looking for a more direct way of pounding the gores into shape. Somebody suggested dynamite.

Down at El Toro Marine base in Orange County they found a 60,000 gallon water tank. They sank the forming die to the bottom of the tank with the flat sheet of aluminum resting on it; then they laid a pattern of explosive above the sheet and hit the detonator. The water erupted, the shock wave hit the aluminum plate, and the plate bent — but not all the way. They found that three successive explosions were needed to do the trick.”

To weld the common bulkhead gores together into domes, North American built another automatic welding machine that would crawl along the seam and weld it, constantly adjusting the weld parameters as the shape and thickness of the aluminum sheets changed.

After this came the “ultimate problem”: turning the formed and welded domes of aluminum into honeycomb panels for the common bulkhead: 

“Though the completed domes were rigid, they still had a tendency to sag, so the lower dome was inflated with air pressure, and, in an intricately timed industrial ballet, sheets of heat-sensitive adhesive were taken from the freezer and stuck to the honeycomb, the precut honeycomb panels were fitted to the surface of the aluminum, and the bond was cured by baking the dome at 300 degrees in the worlds largest pressure cooker. Then the upper dome was lifted by a huge hat-shaped vacuum chamber that sucked it up to the proper curvature, and the measurements were taken underneath to guide the shaping of the face of the honeycomb on the lower dome. For this formidable task they wound up building the world’s largest lathe. As the lower dome rotated, a grinder controlled by data tapes moved slowly up the curving surface and ate the honeycomb down to the measured thickness. After a series of trial fittings and regrindings, the honeycomb was covered with sheets of adhesive, the upper dome was lowered onto the honeycomb, and the whole thing was sent back to the oven. Finally the two aluminum domes, which touched only at their circumference, were welded together.”

Not every effort at weight reduction was solved through clever (if complicated) ideas like cold-strengthened aluminum or the common bulkhead. Much of the effort was achieved by pure brute force: parts would be fabricated, tested until failure, and then redesigned to be slimmer until they broke at exactly the required load (scaled by an appropriate safety factor).

There were similar challenges in building the Apollo spacecraft, with its 2 million parts and 15 miles of electrical wiring. Minimizing weight, achieving Swiss-watch levels of precision, and withstand extreme forces. The heat shield, for instance, needed to protect the spacecraft from temperatures hotter than the surface of the sun when it reentered the atmosphere. Avco, a Delaware company, developed an epoxy resin that could withstand this temperature, charring and flaking off as it was heated (and dumping heat in the process). Turning this epoxy into a functioning heat shield was a challenge: 

“To keep it in place, Avco’s builders bonded fiberglass honeycomb to the capsule surface and pumped the epoxy into each individual cell with a caulking gun. It had to be done by hand; there was no other way. And if X-ray inspection revealed a bubble, they cleaned that cell out with a dental drill and tried again. A square inch at a time, they worked their way around the surface until they had filled all 380,000 holes.”

Even something as seemingly straightforward as the capsule’s parachute took on new heights of difficulty under the program’s extreme requirements. Storage space limitations meant that over an acre of cloth and several miles of parachute cord needed to fit into an area not much larger than a dresser drawer, with a weight of less than 500 pounds. To achieve this, Northrop, the subcontractor for the parachute design, developed a special nylon fabric, and put it under a hydraulic press inside a vacuum chamber, sucking the air out and compressing the fabric until it had the density of a block of wood. When the parachute cord was found to cause problems due to excessive stretching (the capsule would “yo-yo” up and down as it descended, causing instabilities), Northrop created a new type of cord out of fishing line, stitching it into a mesh that wouldn’t stretch.

Figuring out how to meet the program’s requirements inevitably resulted in several blind alleys. North American spent untold amounts of time trying to get bonded panels of insulation to work before they developed their spray-on method. To try and build a fuel gauge that would work in zero gravity, North American struggled for months with a “nuclear fuel gauge” that sent a small amount of radiation through the tank and measured how much it was attenuated by the fuel (it was eventually scrapped in favor of a simpler electrical gauge).

While Storms’ Space division struggled with the Apollo spacecraft and the S-II, the Rocketdyne division was having its own difficulties with the Saturn V engines, chief among them combustion instability on the F-1 engine. Combustion instability — getting the fuel and oxidizer to burn smoothly and consistently — is, according to former Apollo engineer Tom Kelly, “the most fundamental technical problem in a liquid-propellant rocket engine," and had plagued rocket builders since von Braun first started launching rockets at Peenemünde. Most difficult of all was that combustion instability wouldn’t occur consistently: a rocket engine might work fine on many firings, only to exhibit catastrophic instability on the next. It had traditionally been addressed with “cut and try” methods, or making changes in the geometry of various engine parts until the problem no longer appeared, which didn’t give many clues for how to solve it on a new type of engine.

Combustion instability destroyed an F-1 engine during testing in 1962 and two more in 1963, threatening to derail the whole Apollo program if not addressed. Rocketdyne engineers eventually concluded that there was no way to prevent an engine as large as the F-1 from becoming unstable. Instead, they would aim for “dynamic stability”: ensuring that any instabilities that did occur would quickly damp themselves out.

To study instability in the engine, Rocketdyne engineers took extreme measures. They built a bomb and placed it in the nozzle of the engine, creating combustion instability upon detonation. Armed with the ability to create a combustion instability whenever they wanted, Rocketdyne and NASA engineers were eventually able to achieve dynamic stability by refining the design of the baffles within the engine.

The Rocketdyne engine used on the upper stages, the liquid-hydrogen burning J-2, had its own set of difficulties. Preventing hydrogen leaks proved difficult, and ultimately special welded joints and joint seals had to be developed. New low-temperature lubrication systems had to be invented. With liquid oxygen, the low temperature caused frost to accumulate on components, which actually served to insulate and protect them. The frost also served to damp any vibrations that occurred. But liquid hydrogen was so cold that air touching the components that carried it would liquify rather than frost over, dramatically increasing heat leakage rather than insulating them and making the parts susceptible to vibration damage. This ultimately required special vacuum-jacketed piping to be developed to prevent heat leaks and redesigned fuel lines to prevent vibration failures.

Angle of Attack focuses on the difficulties of North American Aviation and its subcontractors, but they were far from the only contractors having trouble. On the third stage, Douglas Aircraft was having its own challenges welding. Like North American, Douglas built an automatic machine for welding aluminum sheets together, which initially wandered all over the place instead of following the weld seam. The problem was eventually traced to the seam tracking system, which worked by detecting changes in electrical current across the seam. The machining standards on the aluminum sheets were so high that when placed next to each other, there was no detectable gap: the machine just saw a continuous sheet of aluminum (this was eventually fixed by redesigning the machine to be more sensitive). For tank insulation, Douglas engineers initially wanted to use balsa wood, but realized there wasn’t enough balsa in the entire world to do the job, forcing them to develop a “synthetic” balsa. The third stage also had its own common bulkhead built from aluminum honeycomb, which required, among other things, sanding down the surface of the honeycomb by hand to get it to the proper dimensions.

On the first stage, which was built by Boeing, the challenges were primarily due to the immense size of the rocket. It required the largest aluminum forgings that had ever been made in the US, produced in one of the country’s two 50,000 ton presses that could handle such large components. Because even a fingerprint could cause catastrophic contamination, elaborate cleaning procedures were developed, including putting parts in “the world’s largest dishwasher”: a box 40 feet on a side and 20 feet tall, inside of which a revolving pipe mechanism would spray everything down with special cleaning chemicals. Like North American and Douglas, Boeing also had to do its welding in precisely controlled, clean-room environments. Welding was done by welding teams of 10 to 15 members, using carefully designed procedures sequenced “like the countdown of a launch vehicle," and could last up to 8 hours.

Many of the difficulties building Apollo stemmed from the immense reliability requirements. Components had to both survive the immense forces of launch and reentry, while being as light as possible. Even the smallest failure in a component could be disastrous. 

To minimize the chance of failure, a system called “traceability” was invented: tracing the exact chemical makeup and manufacturing steps that went into producing every single part, all the way back to the mine or tree that produced it. At one point, traceability caused a minor scandal when North American was forced to explain why the bolts it used were so much more expensive than the 59 cent bolts that could be picked up at a hardware store:

“Charlie Feltz…explained to the committee in his quaint, sunbleached Texas style that there were eleven steps in the manufacture of these bolts and they had to be certified at every step. Not only had the bolt itself been subjected to rigorous testing, but the steel rod it was milled from had been tested, as had the billet from which the rod was extruded and the ingot from which the billet was forged. Indeed, they knew where the iron ore had come from — the Mesabi Range north of Duluth — and they knew which mine and what shaft. And when you factored in all that extra rigmarole, said Charlie, it turned out the actual cost of the damn bolts was not $8 or $9 but more like $32.”

As parts were assembled into finished rockets and spacecraft, there were even more tests. Tests of tank pressurization, of rocket firing, of launch escape systems, of piping systems, of lander impact. Tests on every component, sub assembly and assembly on the rocket, in every condition it might face. Though the spacecraft would spend just 0.1% of its time in the earth’s atmosphere, it underwent 11,000 hours of wind tunnel testing, using 37 different models of the ship. By some estimates, half the cost of the Apollo program went into testing. And whenever a test was failed (and there were, of course, numerous test failures), the chain of events that had caused it had to be traced back and the problem eliminated.

At one point, for instance, a titanium fuel tank for the Apollo spacecraft exploded during pressure testing, followed by another tank explosion a few days later. North American investigated, and found that identical tanks at the manufacturer’s plant in Indianapolis had worked just fine. North American eventually traced the problem to the nitrogen tetroxide the tanks had been filled with during the test. The refinery that produced it had changed its process to make it as pure as possible, and it turned out that when nitrogen tetroxide is purer than 99% it would eat away the titanium. A tiny amount of water added to bring down its purity corrected the problem.

And while contractors were solving manufacturing problems and trying to pass stringent test requirements, the design of what they were building was constantly changing. When North American won its contracts, many of the program specifics were still in flux. It wasn’t until months after North American won the contract for the Apollo spacecraft that a landing method was chosen. Ultimately lunar orbit rendezvous was chosen, using a small landing craft that would dock with the Apollo mothership, but this was just one of several competing options, including landing the entire Apollo spacecraft on the lunar surface. And it wasn’t until 1963 that a water return landing for the Apollo capsule was decided on. As early conceptual sketches were refined into detailed engineering schematics, engineers inevitably reconsidered their assumptions and reworked basic design strategies. And whenever a test required a change, or a subsystem was swapped out, or a part was redesigned to make it lighter, that rippled back, requiring changes to every part that it interacted with. This created a veritable river of blueprint revisions. During the first three years, change-orders for the Apollo spacecraft averaged 1000 per month. Part designs were obsolete by the time they hit the factory floor.

Under these conditions, it's inevitable that the contractors would fall behind schedule, but North American struggled most of all. For one, the technical problems it needed to solve were among the most difficult of the entire program. Boeing had difficulties fabricating the immense first stage, but it didn’t need to trim weight to the extent that North American did, or handle liquid hydrogen fuel (which had only existed in tiny quantities in labs prior to the Saturn I and V). Douglas Aircraft likewise seemed to have things easier, in part because the third stage was an adaptation of the Saturn I second stage. Though Douglas also used a common bulkhead for its third stage, it was smaller, and used a heavier but easier-to-fabricate dome rather than the complex ellipsoid of the North American common bulkhead. Douglas also didn’t need to solve the complex problems of bonding the insulation to the outside of the tank rather than the inside.

North American also struggled because its items were on the critical path for meeting the launch date. This meant it received much more pressure from NASA, which criticized it for delays and problems that went relatively unnoticed at other manufacturers. Grumman (which merged with Northrop in 1994), for instance, was nearly a year behind schedule in producing the lunar module, and only caught up because of the catastrophic Apollo 1 fire that delayed the entire program. Boeing was likewise behind by three months on the first stage by 1964. Among other contractors, North American’s second stage became known as the “umbrella," as it protected them from NASA’s ire.

As they struggled to meet the program schedule, NASA began to see North American as a threat to the entire program. During the first years of the program, notes Charles Murray in “Apollo," North American acquired a “very bad reputation around NASA."⁶ After a completed second stage was destroyed during pressure testing in 1965, further delaying the program and driving up the already over-budget cost, NASA sent a “Tiger Team” headed by General Sam Phillips to investigate North American’s Space division.

Phillips’ report was scathing. He stated that “I am definitely not satisfied with the progress and outlook of either program…I could not find a substantive basis for confidence in future performance," and provided a long list of things that needed to change. In private, Philips was even harsher, writing in a memo that “my people and I have completely lost confidence in the ability of North American’s competence as an organization…and I seriously question whether there is any sincere intent and determination by North American to do this job properly.” The memo ended with a list of people that should be fired, which included Storms and most of the rest of top management. Eberhard Rees, a Germany V-2 engineer who was now von Braun’s deputy on the Saturn program, likewise wrote in a private memo that NASA needed to “resort to very drastic measures," up to finding a new contractor. “For me, it is just unbearable to deal further with a non-performing contractor who has the government ‘tightly over a barrel’ when it comes to a multibillion-dollar venture of such national importance," wrote Rees.

To get the programs into shape, Storms rearranged leadership. Former Air Force general Robert Greer was placed in charge of the second stage program, and reorganized North American’s efforts to “stop the project from choking on its own complexity." Storms put some of North American’s best talent, including director of engineering George Jeffs, on the Apollo spacecraft program to get it under control. Slowly, these efforts bore fruit, and a few months later NASA administrator George Mueller (General Phillips boss) wrote to North American that “your recent efforts to improve stage schedule position have been most gratifying." By 1966, successful unmanned test flights of the Apollo spacecraft (using a Saturn I rocket) were taking place.

The fall of Harrison Storms

The ending of the Apollo program story is a happy one. America successfully lands a man on the moon, a feat that no other country has managed to achieve. The story of North American and Harrison Storms is not so happy. 

In a test for the first crewed Apollo flight in January 1967, a catastrophic fire broke out in the capsule. The capsule was destroyed, and astronauts Gus Grissom, Ed White, and Roger Chaffee were killed.

The disaster resulted in a congressional investigation of NASA and the Apollo program, and NASA convened an Apollo Review Board to investigate the cause of the fire. As the investigations proceeded, it became clear that North American was being positioned to take the blame for the fire. The Senate Space Committee dug up the Philips report detailing numerous problems with North American’s management of the spacecraft and second stage. It also uncovered the fact that when NASA selected a contractor for the spacecraft, North American had in fact lost to Martin on the technical evaluation. Storms got word that the head of NASA’s Apollo Review Board had orders “to clobber you guys." When NASA’s 2,300 page report was released, it didn’t point to a definitive cause of the fire, but suggestively pointed to numerous areas of “poor installation, design, and workmanship” in the wiring and environmental control systems, placing the blame squarely with North American.

Within North American, there was debate on how to respond. Regardless of what caused the spark that ignited the fire (and none of NASA’s specific examples of supposed poor workmanship were thought to be related, even by NASA), the reason the fire was so catastrophic was obvious to everyone: the atmosphere of pure oxygen within the capsule. NASA had decided to use pure oxygen against North American’s strenuous objections, as it simplified the environmental control system (only one gas instead of a complex mixture of gasses). Charlie Feltz had told NASA that “it’s the wrong thing to do," to which NASA had responded “You’re the contractor, you do as you’re told. Period."

What’s more, the level of oxygen in the capsule during the fire was more than three times as high as what North American had designed for: 16 pounds per square inch vs. 5 pounds in the design spec. This high level of oxygen had been used, without North American’s knowledge, to ensure the capsule stayed pressurized relative to the outside atmosphere while it was still on the ground. In such an environment, even the smallest spark would burn rapidly and fiercely:

“At five pounds — the pressure inside the command module in orbit — a lighted cigarette would merely burn rapidly; at sixteen pounds, the cigarette would vanish in a flash along with all your hair and your clothes as well.”

On top of this, the hatch for the spacecraft had been designed to swing inward, which took nearly 90 seconds to open even in the best case scenario. North American had favored an outward-swinging hatch that could be rapidly ejected by explosive bolts, but this had been rejected by NASA, due to concerns over explosive bolts accidentally getting triggered.

Going over the project documents, North American management saw more than enough evidence to “sink NASA." North American had strenuously voiced its concerns about the risks of a pure oxygen environment and an inward-swinging hatch in writing numerous times during the program. In fact, Storms had refused to accept the requirement for a pure-oxygen environment until it was officially documented and the contract was amended.

But of course North American didn’t want to sink NASA: it wanted to continue selling rockets and spacecraft, and it couldn’t do that if NASA got its wings clipped. North American management eventually decided that it could weather the storm more easily than NASA, and that NASA (in gratitude) might reward it with future business (this last guess turned out to be prescient, as North American would win the bid to build the Space Shuttle a few years later).

But taking the blame meant heads needed to roll. NASA had already offered up a sacrifice in the form of Apollo manager Joe Shea, and NASA insisted that Harrison Storms be fired. North American complied, shuffling Storms to a paper-pushing job elsewhere in the company. He was replaced with John Bergen, an executive from Martin, the company Storms had beat to win the Apollo spacecraft project in the first place. The day Storms leaves was a dark day for North American:

“The secretaries were in tears, so were the telephone operators, and as the word filtered out into the plant, so were some of the riveters and welders. He had lifted them out of the humdrum of their ordinary lives and put them to work on one of the greatest adventures in history, and now a bunch of sonsabitches who probably couldn’t find their asses with both hands were yanking him out of the saddle just short of the finish line. As the news flashed out through the far-flung division to the 35,000 people in a hundred locations to whom he was known simply as Stormy, anger welled up like the sea. It was outrageous. If Harrison Storms hadn’t held everybody’s feet to the fire on the S-2 common bulkhead, there would be no moon landing in this decade; there was no way the Saturn 5 could have lifted the weight of the other design. And the spacecraft itself was unquestionably a masterwork — a labyrinth of systems more complicated than an aircraft carrier packed into a stainless-steel phone booth — and anybody with hands-on experience knew that it was the finest piece of machinery ever assembled. The bastards should have been carrying Stormy around on their shoulders instead of tarring him with this terrible brush.”

Harrison Storms retired three years later, though he would continue to work for North American (as well as other aerospace companies) as a consultant. He never again did any notable aerospace work, and died in 1992 in relative obscurity, a few months before Angle of Attack was published.

North American’s story isn’t particularly happy either. Shortly after the Apollo 1 fire, it was acquired by Rockwell, becoming North American Rockwell. Though the combined company successfully supplied the spacecraft and rocket stages for every Apollo mission, and would go on to build the Space Shuttles, Rockwell management would gradually push the company away from aerospace work, which was “not responsible to the Harvard Business School style of quarterly management." By 1991, the company wouldn’t have a single airplane contract on the books. In 2001, the combined company (by then known as Rockwell International) would be broken up for parts, and today as far as I can tell the only corporate remnant of North American is Rocketdyne, which after being sold a few times eventually became a part of Aerojet.

The possibilities of Apollo

Reading about the enormous efforts it took to get the Apollo rockets and spacecraft built, I was struck by how contingent the whole program seemed. Despite being one of the crowning achievements of civilization, Apollo might not have happened at all if any number of things had gone slightly differently.

For one, though the Apollo program required major advances in not only aerospace but nearly every field of engineering, it was only possible because of the previous two decades of enormous government investments in aerospace and rocket technology, driven by the exigencies of WWII and the Cold War. NASA, in fact, tried very hard to use existing technology wherever possible during the program, preferring the tried and true rather than the novel and unknown. In North American’s case, it was its experience building the Navaho, the X-15, and the B-70, along with the knowledge of titanium created by Lockheed’s A-12 program, that created the capabilities needed to build Apollo.

WWII likewise probably created the necessary army of aerospace and rocket experts who could get things done by pure force of will. Harrison Storms is, of course, an example. At one point in the book Storms refused to allow a missile test to be canceled after a system anomaly:

“Storms said he wanted to talk to the flight crew. He grabbed the microphone and he spoke to the pilot high over the Atlantic. “How much gas do you have?” After a moment of puzzled silence, the pilot said he had about five hours of fuel on board. “Then turn back to the Cape and go back to the top of the checklist and start over,” said Storms. The people in the room looked at each other — and so, no doubt, did the men in the cockpit 3000 miles away. The test director said, “Mr. Storms, we can’t restart the test. The Air Force has closed the test range.” “Open it.” “But everybody’s gone home!” “Get ‘em back,” said Storms. “And don’t land that bomber with more than a teacup of gas in the tank.” ...two hours later, on the second pass through the checklist, somebody spotted a switch he’d forgotten to close the first time. They launched the missile and it flew like an arrow to the target.”

But everyone of note in the Apollo program seems to show the same basic profile. Wernher von Braun, for instance, managed to build his rocketry organization in Germany, move the entire thing across a war-torn country to get captured by the Americans, and rebuild it in the homeland of his former enemy in his quest to put rockets into space. Without this talent pool of hard-charging aerospace experts, Apollo probably also doesn’t happen, at least on the timeline that it did.

And that timeline is probably fairly sensitive. It couldn’t have happened much earlier, as rocket and aerospace technology wasn’t yet good enough. But it probably couldn’t have happened much later either. Even while Apollo progressed, we see a shift in the tenor of politics, marked by things like the National Environmental Policy Act. The era of the government technocrat given free reign to run projects as they saw fit was gradually replaced by decreasing trust in government and increased oversight, bureaucracy, administrative restrictions and risk aversion. Combined with a decade of energy crises, it's hard to imagine an Apollo program that started in 1971 as being successful.

And the 1961 program could have easily been derailed. As early as 1963, people were already tired of the enormous amount of money that the Apollo program was absorbing (it would ultimately cost over $250 billion in 2020 dollars). Kennedy’s assassination galvanized the program and gave it the momentum to continue forward, and “insulate[d] Apollo from the horrors yet to come”:

“A week before Dallas, the Senate had cut $600 million from Kennedy’s NASA budget; a week later, it named the Florida launch complex after him and put the money back on the table. From then on, as one engineer said, ‘Apollo had only three sacred specifications; Man. Moon. Decade.’”

Angle of Attack: Harrison Storms and the Race to the Moon is available on Amazon.

1. “The technology that goes into the B-70," Rees argues, “is the technology that must go into space vehicles when they’re built.

2. This was, in part, due to the Treaty of Versailles, which forbade Germany from developing long-ranged artillery. When Hitler came to power in 1937, he thus began to put effort into developing rocket technology, which the Treaty did not cover.

3. V-2’s were also built using slave labor.

4. Liquid hydrogen has a higher specific impulse than kerosene, producing more thrust per unit weight.

5. In “Apollo," Charles Murray describes this somewhat differently. According to him, bidding on the Apollo spacecraft as almost an afterthought by North American, though this doesn’t square with the immense amount of money Storms spent on the bid.

6. Yes, that Charles Murray.