Summer–Fall 1968
Once NASA decided to go forward with a journey to the moon before the year was even out, the Apollo 8 astronauts, mission controllers, and flight planners had to move fast. They had only sixteen weeks before launch day, and now the clock hands counting off that time began moving very quickly. Wernher von Braun, as much as anyone, was quite open about how urgent the job was, and he found an evocative way to express it, even though English was not his first language.
“The moon in the sky,” he would say, “has become a deadline display device.” And unlike the countdown clocks at Cape Kennedy, it was a device that nobody could ever turn off.
For von Braun, the sixteen weeks would have to be front-loaded, because although the Saturn V would play the most short-lived role in the lunar enterprise—less than three hours after liftoff, the rocket’s 363-foot stack would be reduced to just 35 feet of command and service module—it was decidedly the most difficult component to prepare and ship and configure for launch. And this time, the Saturn had to fly true. If it went pogo again, it might shake the astronauts to death. If it failed catastrophically, it would unleash a monstrous and inevitably lethal explosion.
Immediately following the disastrous flight of Apollo 6, von Braun had gathered his entire Huntsville team to fix the problems that had almost wrecked the ship. Just to make sure he had all of the smartest heads together in the same room, he had also summoned—either in person or by phone—nearly one thousand other engineers from every contractor, subcontractor, and sub-subcontractor who might have a useful thought about how best to sort out the giant machine’s multiple defects.
The pogo problem, the most critical issue, was traced to a simple fact that is true of all rockets: they burn fuel, and as that fuel vanishes, it leaves nothing behind. The lower the level of kerosene, liquid hydrogen, and liquid oxygen in the tanks, the less mass there was to absorb the rocket’s vibrations, which, in turn, had led to the pogo oscillations.
The answer was to keep that emptying space full—but with what? The Saturn V’s engines had only one speed, which was full speed, and the only way for the rocket to accelerate continually as it climbed was for it to get lighter and lighter as its fuel burned away, resulting in a fixed thrust that was propelling an ever-dwindling mass. That’s how you got from zero miles per hour to the nearly 25,000 miles per hour it would take to get to the moon. The problem was, if you added something to the tanks to fill the empty spaces, you would be adding weight, too.
The engineers solved the puzzle neatly with helium. It was inert and it was light; in fact, at atmospheric pressure, the stuff actually had negative weight. That would not be the case in the Saturn V, since the helium would be pressurized and thus more densely packed, but it still weighed effectively nothing. All you had to do was steadily pump helium into the tops of the tanks to take the place of the fuel as it burned away, and the gas would absorb the launch oscillations like a properly inflated tire cushioning the bumps in a road.
The failure of the two engines on the second stage was a knottier issue. Pogoing was part of the problem here, too: during the flight of Apollo 6, it was so severe that it had caused the I-beam deflection. But another part was simple sloppiness—potentially deadly sloppiness—in the factory, where the electrical leads to the two engines had been crossed. The wrong power systems were thus feeding the engines, meaning the systems couldn’t respond properly when one engine or the other called for more or less power. That boneheaded error was quickly exposed in the post–Apollo 6 investigation, and von Braun had been assured that there would be no such nonsense going forward.
These and other fixes and refinements gave von Braun confidence that the next Saturn V would be fit to carry a crew. And this made him confident that the Apollo 8 crew could go to the moon.
“Once you put a man on it, it doesn’t matter where you send it,” he said in the days after Gilruth, Low, and Kraft came to visit him in Huntsville.
NASA itself wasn’t yet convinced, however. Before the agency would approve von Braun’s rocket for a manned flight, the chief designer would have to make his case to Dieter Grau.
Like von Braun, Grau had been one of the German rocket engineers who’d come to the States after the war. He had worked for von Braun in Germany, and now he worked for him in Alabama as well. But Grau was also Huntsville’s chief of quality and reliability operations, and in that capacity he would, at certain moments, be his boss’s boss. This was one of them: von Braun could argue all he wanted that his rocket was ready for flight, but it wasn’t going anywhere until Grau gave it the okay.
“What more should be done?” von Braun asked him in September, after he had done all the work he could do on the rocket.
“I want the opportunity to do one more complete check,” Grau answered.
Von Braun agreed, and over the course of the next few weeks, the quality and reliability team went over the schematics and design history of every element of the Saturn V that would carry Apollo 8, as well as the performance telemetry from Apollos 4 and 6 and the contractors’ records of the testing they had done after those earlier flights. Precisely as Grau had feared, several little problems presented themselves—none critical, perhaps, but none acceptable. Only when every component in every system in the rocket had been set right and certified fit did Grau give the machine his approval.
Even then it was a conditional approval. Once the rocket was on the launchpad at Cape Kennedy, it would be out of Huntsville’s hands, but there was plenty of proxy hardware still in Alabama. Grau ordered up continued firing tests on Saturn V engines straight through the fall. The final tests would take place in December, with the last one scheduled for the eighteenth of that month, just three days before Apollo 8’s launch.
* * *
Creating the necessary software and writing the needed computer code was another hurdle for the Apollo 8 team. Gene Kranz might have given Bob Ernal the use of every computer in Buildings 12 and 30 for an entire weekend, but a thumbs-up from Ernal after just two days of work was hardly the final answer. It was, in fact, only the signal that it was time for a lot more people to get together and do a lot more work. For this, Bill Tindall would again play a role.
Tindall, who had played a central part in the debate by Chris Kraft’s team about whether Apollo 8 should orbit the moon, was known by Gene Kranz and other higher-ups at NASA as “the architect”—a hat tip to his status as the person who had blueprinted all of the Apollo’s computer programs. To his subordinates, he was known for his blunt, sometimes blistering technical memos, which were unsparing when the architect wanted to make it very clear just what he thought of this or that piece of software built by one of his team members.
As August turned to September, with the sixteen-week clock ticking, Tindall took to spending more and more of his time in Cambridge, Massachusetts, at MIT’s Draper Lab, where the mission’s software was being written and tested. Before long, four hundred engineers were working at MIT under the nominal authority of the university but the practical authority of Tindall. The spacecraft’s memory was divided into thirty-eight separate banks, each capable of storing a full kilobyte of information—or one thousand discrete units of data. That was a lot of computing power to pack into a small space.
The focus of most of their work was the Colossus software, the system that—after winning a years-long competition with other programs—had been chosen as the one that would be used to navigate to and from the moon. But Colossus wasn’t completely ready, and while the pace of the work to make it fit for flight had already been fierce, now it would have to move at a full sprint.
The most serious concern was what to do in the event of software crashes, which occurred often and might never be entirely eliminated from the program. The key to responding to crashes was how quickly and seamlessly the system could reboot. On a three-day outward coast to the moon, you might be able to tolerate it if your computer went off-line for a little while, but in the critical minutes or seconds before the engine burn that would cause the spacecraft to enter lunar orbit, a software crash would be disastrous.
The work at MIT was thus focused on ensuring that the restarts following a crash would be fast and automatic. The computer’s core operating program would have to know which systems to start up in which sequence, how to check the soundness of each system as it came online, and how to suspend all other nonessential operations aboard the spacecraft while that work was being done.
Tindall drove the team hard and was not above using the engineers’ own, often grand, egos against them. If MIT couldn’t handle the job, he would casually muse, he could always call in some hotshots from IBM for consultations. The newly motivated MIT engineers would promptly assure him that no such help would be necessary.
As the fall wore on, the bugs got written out of the software. When Colossus was finally considered ready for flight, nobody could say with certainty if that confidence was well placed. But Tindall pronounced himself happy, and that counted for a lot.
* * *
Assuming the Apollo 8 spacecraft did get off the ground and did make it to the moon and did get back to Earth safely, there was still the business of rescuing the astronauts after splashdown. Technically, it wasn’t a difficult problem: you sent out ships and plucked the crew out of the water, and in the seven years Americans had been flying in space, NASA and the Navy had gotten very good at that job. Nevertheless, it was always a massive logistical headache.
There were a few possible launch windows at the end of 1968, and all of them were dictated by the relative positions of the Earth and the moon. The moon was a moving target, orbiting the Earth at a speed of 2,288 miles per hour. This meant that when Apollo 8 left the ground, it would be aimed not at the spot the moon was occupying in the sky at that moment but toward where it would be three days hence—much the way hunters direct their buckshot ahead of a flock of flying ducks, rather than at the ducks themselves.
For Apollo 8 to succeed, three conditions would have to be met. First, Earth would have to be at the precise spot in its rotation where the Saturn V could leave the ground, enter Earth orbit, and then blast out to the moon at the proper angle of approach. Second, either the Atlantic or the Pacific Ocean would have to rotate into position so that it would be underneath the spacecraft when, six days after launch, it arced through the atmosphere for splashdown. Third, the moon would have to be in the right spot in its waxing and waning phases so that the portions of the surface the crew planned to survey as future landing sites would be properly illuminated. After much calculation, NASA determined that the best of the possible windows put the liftoff at 7:51 a.m. eastern time on December 21, with the spacecraft going into lunar orbit on Christmas Eve and splashing down in the Pacific Ocean southwest of Hawaii in the predawn hours of December 27.
The location and timing of Apollo 8’s splashdown required that a complex set of orders be given to the recovery ships, and some of those ships might already have commitments. The Navy was planning to give its sailors a short Christmas vacation, and the decision to keep some of them working so that they could recover the spacecraft would rest with Admiral John McCain Jr., the commander of the Pacific Fleet and of all naval forces in Vietnam. McCain had a lot of skin in the Vietnam mess; his firstborn son, naval aviator John McCain III, was being held as a prisoner of war in North Vietnam, having been shot down and captured almost a year earlier. Still, the Navy and its senior officers had myriad responsibilities, and they included supporting the nation’s space program—as long as that support didn’t compromise naval operations.
It was Chris Kraft’s job to persuade Admiral McCain that Apollo 8 met these criteria. In October, Kraft flew out to Hawaii and addressed an auditorium full of Navy brass, though most of his attention was directed at McCain, who sat smoking a large cigar, surrounded by his officers. Point by point, the space man delivered a meticulous presentation to the Navy man about the mission’s goals, its stakes, and, most important, its vaulting ambition: getting American astronauts out to the moon and back before the Russians could beat them to it.
“Admiral,” Kraft concluded, “I realize that the Navy has made its Christmas plans, and I’m asking you to change them. I’m here to request that the Navy support us and have ships out there before we launch and through Christmas. We need you.”
McCain did not need to give the request more than a moment’s reflection. “Best damn briefing I’ve ever had,” he said. “Give this young man anything he wants.”
* * *
If the teams in Huntsville and Cambridge and NASA headquarters were sprinting toward December, the men in Houston were working even harder—specifically, the three who would fly the Apollo 8 spacecraft and the dozens who would rotate through the consoles at Mission Control. And if the engineers at MIT had their Tindall to run them ragged, the people in Houston had their simsups, and they were even worse. The simsups were the simulation supervisors, the men whose job it was to make the lives of a lot of other people miserable. You could actually see the simsups at work; they sat at their own bank of consoles on the right side of the Mission Control auditorium. But you couldn’t quite get to them: a wall of glass separated them from the rest of the room, which was probably just as well.
For both the astronauts and the mission controllers, the bulk of the training involved running simulated missions, then running them again and again and again so that everyone knew every step in every possible flight plan deeply, exhaustively, reflexively. And then the simsups would blow everything up. They would allow a routine rehearsal in Mission Control to run for a while, and then, with no warning, they would shut down three of the imaginary first-stage Saturn V engines when the rocket was only a thousand feet off the launchpad. Or they would kill the communications systems five minutes after the crew had left Earth orbit on the way to the moon, and when controllers would try to switch to the backup system, they would take that out, too. Or they would crash the spacecraft’s environmental system and then order the man at the environmental control console to get the system configured again before the astronauts died from hypothermia or lack of oxygen.
The astronauts rehearsed in a spacecraft simulator elsewhere on the Houston campus, and the simsups were just as merciless when working with them. They would get the crew all the way to the far side of the moon, overburn their main engine, and then give them precisely three minutes to sort out the problem before their orbit irreversibly decayed and they headed for a crash landing on the lunar surface. They would send the command module into a high-speed spin halfway to the moon and then kill the thruster controls, meaning the crew would have to bring their backup systems online before they could even begin to stabilize their spacecraft—and they would have to do it all before the simulated spin rate reached sixty revolutions per minute, or one per second, at which point real astronauts in a spacecraft that was actually spinning would suffer extreme vertigo and lose consciousness.
Sometimes the simulations were run only with the astronauts or only with the mission controllers. Other times the simsups would conduct so-called integrated sims, when the people on both ends of the voice and telemetry links would play the same roles they’d take days or weeks later when the mission actually flew. That was the way astronauts and mission controllers had trained for every flight that had ever flown, and that was the way they trained for Apollo 8. Except this time they were training to fly to the moon—or, more accurately, they were learning to fly to the moon, since no one had ever done it before.
* * *
Throughout the run-up to the launch of Apollo 8, it had remained an absolutely inviolable condition that Apollo 7’s mission should, in George Low’s words, be “very good, if not perfect,” or the Apollo 8 crew would not be going to the moon.
Apollo 7’s commander was Wally Schirra, and the happy, jokey Wally known to all during the Gemini era was still in little evidence. Ever since the fire, he had been replaced by the Wally who prowled and scowled on North American Aviation’s factory floor, determined to prevent another disaster such as the one that killed three of his fellow astronauts. Not long before the launch of Apollo 7, he and his crew—Walt Cunningham and Donn Eisele—sat for a press conference in Houston, and from the beginning, Schirra seemed testy, impatient, and not a bit happy to be there. Things got especially awkward when someone asked him how comfortable he was with the soundness of his spacecraft.
“We’ve basically lived with Apollo 7 at the plant, we’ve lived with it at the Cape, and if somebody takes even a small component off it, we immediately become furious and say, ‘Why did you remove it?’” he responded. “We expect answers immediately.”
The journalists in the room exchanged sidelong glances, while the North American executives at the meeting shifted uncomfortably. This kind of talk was always awkward, although tolerable behind closed doors at the factory. But to share it with the press? It just wasn’t done.
And Wally wasn’t finished.
“In fact,” he went on, “I’m waiting for an answer, as an example, why someone took our hatch cover off and took it to Downey and we haven’t got the answer on why.”
The North American men blanched. Not the hatch, anything but the hatch—the bank-vault door that had killed Grissom, White, and Chaffee. If the engineers had removed the hatch from the spacecraft at the Cape and flown it back to California for adjustments, they’d probably had a good reason. But the mere mention of that one particular part sent a shudder through the room.
“Pardon me, I just went into shock,” said Eisele, a rookie who had no business speaking out of turn. But as long as he was following the lead of the commander, he apparently believed, he could say anything he wanted.
“I know,” said Schirra.
“Oh boy,” answered Eisele.
At 11:02 a.m. on October 11, Apollo 7 blasted off from Cape Kennedy atop a Saturn 1B rocket, the smaller version of the Saturn that would be used for this first manned mission of the Apollo series. The booster worked flawlessly, and in the early going the spacecraft also performed well. But the crew was another matter, and the three men who had been trouble on the ground became unbearable in space. Just fourteen hours after they took off, Cunningham reported to the ground that Schirra had come down with a serious head cold. In the cramped command module, where the air was recirculated, the windows couldn’t be opened, and any surface touched by one man would inevitably be touched by the others, Cunningham and Eisele quickly caught the bug, too. The crew members’ bad health fouled their tempers further, and it didn’t help that the people who had drawn up the flight plan—itchy after nearly two years without a manned space flight—had stuffed the schedule with so many experiments and maneuvers that the astronauts barely had a chance to rest and catch their hacking, congested breath. Before long, barely a pleasant word passed between the crew in orbit and the controllers on the ground.
“I wish you would find out the idiot’s name who thought up this test,” Schirra snapped at the capcom after a navigational exercise did not work as planned. “I want to talk to him personally when I get down.”
“While you’re at it, find out who dreamed up the horizon test, too,” Eisele tossed in. “That was another beauty.”
When a backup evaporator failed to work properly, potentially limiting the astronauts’ water supply, engineers on the ground came up with a makeshift fix—just the kind of work-around technicians were good at devising and that a crew usually appreciated. But not this crew.
“Is this something that somebody’s dreamed up after all these months?” Cunningham growled. “I’ve been told you can’t reservice a secondary evaporator.”
The ground responded that that used to be correct, but now there was a way, and it would be a relatively easy four-step fix.
“Okay, hit me with it,” Cunningham answered. “It looks pretty Mickey Mouse to me, but I’ll stand by if I have to do it.”
Finally, mere insubordination turned to rank mutiny. On October 22, the mission’s final day, it came time for the crew to don their pressure suits and helmets before reentry. This protocol had always been followed and always would be, as far as NASA safety engineers were concerned, in case the fiery plunge through the atmosphere caused a breach in the ship, leading to a sudden depressurization. But Schirra was above such things and refused to wear his helmet or to insist that his crew do so, lest their congested ears suffer from the pressure.
The capcom demanded that the rule be observed, but Schirra was unmoved. Finally Deke Slayton took the microphone and spoke directly to the spacecraft. This was unprecedented: in order to avoid contradictory commands being sent to the ship, the capcom’s voice was the only one the astronauts were supposed to hear. Even Slayton, however, could not persuade Schirra and his crew to put their helmets on.
“I guess you better be prepared to explain in some detail when you land why you haven’t got them on,” Slayton said, finally relenting. “But it’s your neck, and I hope you don’t break it.”
Schirra didn’t break his neck, and Cunningham and Eisele didn’t break theirs. But there was a price of another kind to pay.
Before he’d flown, Schirra had announced that Apollo 7 would be his third and last mission. But Cunningham and Eisele—two men who had been brave enough to be part of the first crew to fly an unproven machine—had very bright lunar futures. After Apollo 7, however, that was all over. No sooner did they land than Kraft dropped the same hammer on them that he had once dropped on Scott Carpenter. Eisele would never fly for him again; Cunningham almost certainly wouldn’t either, though since, of the three, he had offended the least, his sentence included the possibility of parole. But it was a parole that never came, and before long Cunningham, too, was out of the space flight business.
The Apollo spacecraft itself, however, was very much in business. It was all but universally agreed within NASA that the October mission had turned out to be every bit the success Low had demanded: the command and service module had worked almost perfectly for the entire eleven days the crew was aloft. Especially important were the tests of the main engine, which repeatedly lit and shut down precisely on command. If Apollo 7’s engine performed so well in orbit around the Earth, there was no reason to think Apollo 8’s wouldn’t do the same in orbit around the moon.
The lunar mission was on. Schirra, Cunningham, and Eisele would spend the rest of their lives earthbound, but Borman, Lovell, and Anders were going to the moon.
* * *
Nobody at the Central Research Institute building outside of Moscow, where the brain trust of the Soviet space program spent their days, could spare much of a thought for any space mission the Americans might or might not be planning. They had more important matters to mind at home.
In September, more or less as planned, they had launched their Zond 5 spacecraft and whipped it around the moon with a cargo of turtles and worms and insects, bringing them within twelve hundred miles of the lunar far side. The spacecraft returned to Earth, but its aim was poor and it missed the precise atmospheric corridor needed for the high-speed reentry. It didn’t miss by much, however, and although the ride was rough and the spacecraft landed in the Indian Ocean rather than on the steppes of Kazakhstan, the animals survived.
Making the mostly successful mission sweeter for the Soviets, the American surveillance ship USS McMorris happened to be loitering nearby when the Zond was recovered. Moscow assumed that the nest of spies aboard the ship would surely report back to Washington that the Soviet Union was about to beat them in space once more, this time with the first manned mission to the moon.
But before attempting to achieve that milestone, the Soviets would need to launch at least one more unmanned Zond flight, just to increase their confidence that a cosmonaut could survive the ride the turtles and worms and insects had. The Russian space engineers were sure they had solved the reentry problems; to prove the point, this time they would not only fly their animal passengers around the moon and back but land them a precise 9.9 miles from the launchpad. A cosmonaut who touched down that close to the pad would be practically within walking distance of the launch site barracks.
Zond 6 launched on November 10. Like Zond 5, it swung around the far side of the moon and flew straight back to Earth. Then, as it entered the atmosphere, it became clear that this time it would actually do much better than its earlier Zond brother. The ship performed the reentry maneuver almost flawlessly. It plunged through the atmosphere, building up no more heat and no more g’s than a human passenger could easily handle, and headed straight for a landing at the exact spot that had been selected.
The Zond’s parachute deployed when it was supposed to, and its speed of descent slowed as it was supposed to as well. Then, just 3.3 miles above the ground—after a journey of some 230,000 miles—the ship did something it was absolutely not supposed to do until it had touched the Kazakh soil: it jettisoned its parachute. There was nothing to save it then, nor would there have been a way to save a cosmonaut if one had been on board. After falling to the ground like the dead, multi-ton weight it was, the Zond half-buried itself in the earth.
In the Central Research Institute building the next day, Nikolai Pilyugin, the chief designer of the Zond’s guidance system, gathered his engineers for a dressing-down. “Finally,” Pilyugin shouted, “all the systems activated without a problem and you managed to shoot off the parachute when it was almost on the ground! And you were dreaming that we were about to launch a human being?”
The engineers explained that the problem had been an air leak inside the spacecraft. The leak—caused by a faulty rubber gasket—confused the instruments that were supposed to sense atmospheric pressure and, in turn, feed that information to the system that controlled the parachute so that it would know the precise moment to cut the cords. The problem could be fixed, the engineers assured Pilyugin. But they could not assure him that that fix could be made and the Zond could be test-flown in time to beat the onrushing Americans. And they could not assure him that some other problem might not doom the next Zond.
Equal parts mystified and despairing, Pilyugin could only shake his head and turn his attention to Konstantin Davidovich Bushuyev, his resident expert on flight dynamics.
“Konstantin Davidovich,” he asked, “if you could please tell us, after such a good flight, why did you crash the descent module?”