Apollo: The Race to the Moon (47 page)

Against this backdrop, the SimSups began to prepare the flight control team and the astronauts to land on the moon. Since Mercury, all the sims had become closed-loop, meaning that the computers driving the sims could react to the controllers’ actions, changing the scenario as the sim played out. Occasionally, they ran long-duration sims, where the goal was to exercise the flight controllers in a routine, systematic fashion—give them problems that were not imminently mission threatening or life threatening, and make sure that everyone was fully on top of his system.

But the computers’ increased sophistication was more useful for the other kind of sim, the ones that focused on the four most difficult of the mission phases—launch and entry, ascent and descent—and left controllers damp with sweat.* Now the SimSups could double, triple, and quadruple the load on the controllers. A failure in the EECOM system could be followed by an independent failure in the reaction and control system, and that one could be followed by static in the primary voice link. The point was to load the flight control team with several problems at the same time, to approach the controllers’ absolute capacity to absorb and react to a constantly deteriorating situation. “They always tried to run those in double-time, or close to it,” said one controller, throwing perhaps two dozen problems at a single controller in the fifteen minutes prior to landing. You had to field the problem, snap out a call, and move on to the next problem almost instantaneously, or you were going to be working on an old problem while a new one had caused the LEM to crash. The logic behind this brutal routine was that just such a combination of failures would create a disaster in a real flight.

[* In flight operations for Apollo, the words “descent” and “ascent” always meant specifically descent to and ascent from the lunar surface, never the earth’s surface. Several similar potential confusions of wording had to be resolved during the Apollo missions. The old familiar “go” and “no-go” convention, for example. What would the message “go” mean to a crew that had just landed on the moon? That all systems were “go,” meaning okay? Or that they should “go,” meaning leave? Tindall decreed that “stay” and “no-stay” be substituted while the LEM was on the surface.]

The simulation controllers liked to think of themselves as being imaginative and creative. To the controllers, “devious” and “sneaky” seemed more apt. But the men who devised the sims had to be that way. They could assume that any controller who was promoted into the MOCR was cool and quick or else he wouldn’t have gotten to the front room. There was no point in simulating the obvious malfunctions, because he would swat those down instinctively. A good sim was one in which the nature of the problem was masked, in which there was an obvious answer (wrong), a more subtle answer (still wrong), and a right answer that only the most suspicious and persistent mind would see as a possibility.

The one constraint on the SimSups was that, by unwritten understanding, they could not give the flight control team a problem for which there was no answer. No matter how bad things got, the controllers could always say to themselves that somehow, somewhere, it was in their power to save the lives of the crew. Within that single constraint, the SimSups could subject the controllers to any combination of horrors.

“We would dig into the system and find little things,” recalled Dick Koos, who ran sims from the first Mercury days on into Apollo and was nominated by one veteran controller as “the most devious SimSup.” Koos would flip through schematics of the systems, looking for some obscure little diode, and then work out what would happen if that diode should fail. The more utterly unconnected the diode and the resulting failure, the better. The more work it made for the controllers, the better. “Okay, that’ll keep ’em real busy for a while,” Koos would say when he found a good one, and watch with anticipation from behind the window in the Sim Control Area to see how the controllers reacted.

Just because a failure seemed impossible didn’t mean that the SimSups couldn’t go ahead and make it happen anyway. Thus one ebullient and notoriously inventive SimSup, Bob Holkan, was poring through the procedures books looking for arcane rules, and found one for console failure in the MOCR. The procedure said that if a console failed, the controller at that console was to move over to the next console, and so on down the line, the theory being that the last controller, Retro, would move onto Booster’s console, which was empty after the launch phase. Holkan found that the way to create the “maximum motion,” which was his objective, would be to cause a failure in Guido’s console at the far right-hand side of the Trench. But Holkan couldn’t figure out how to make it fail; not surprisingly, the designers of the consoles had gone to great lengths to prevent such an event from occurring.

Holkan found that the only way to shut down a console was to throw its circuit breaker. So Holkan got hold of one of the MOCR’s maintenance men and arranged with him to come in early one morning. Together, they pulled up some of the floor panels, ran a nylon cord from the S.C.A. up into the back of the Guido console, and wrapped the end of the cord around the circuit breaker. They replaced the floor panels, hiding the cord, and Holkan retired to the S.C.A. to wait.

Steve Bales was Guido that morning, and they were running a simulation of the ascent from the moon. At the right moment—which is to say, at the most delicate part of the ascent phase—Holkan pulled the string. “And goddang, if it didn’t work.”

“I heard this big pop,” Bales recalled, “like somebody shot a gun in front of me, and the whole console went blank.” He had to do something—he was sitting in a critical chair for that part of the mission—“but I couldn’t. I froze.” He looked at Jay Greene, the FIDO who was sitting next to him. Greene shrugged; he was busy with his own problems. “I’ve lost my console,” Bales told Flight forlornly, and then just sat there for about thirty seconds, frantically casting about in his mind for something to do. He had never encountered the obscure musical-chairs rule that Holkan had found, but the SimSup always left you a way out, he told himself. He finally thought to ask the O&P controller to put the Guido screen up on one of the 10-foot displays at the front of the MOCR—not the prescribed solution, but a fine demonstration of controller ingenuity. Holkan smiled at the memory. “We went to great lengths to put malfunctions in.”

Usually these were “integrated sims,” in which the astronauts would work from their full-scale simulators in Building 5 or at the Cape, and the flight control team would interact with the crew just as they would during a real flight. By the time of the Apollo flights, the simulators for the flight hardware—both the command module and the LEM—had become highly realistic. The astronauts not only had working versions of all the spacecraft’s switches and levers, they had a view out the window of the LEM simulator showing the lunar surface as it would appear during the descent phase. In another section of Building 5, M.S.C. technicians had built a plaster-of-Paris model of the lunar landing site. As the LEM simulator descended, a camera mounted above the model of the landscape would mimic its movements. Occasionally, this became gulpingly realistic. There was, for example, the time when the sim scenario called for a radar altimeter failure: The astronauts descending in the LEM were led to believe that they were at a ridiculously high altitude for that point in the descent. Failing to recognize the hardware breakdown, the crew accelerated their descent so much that they were unable to recover when, in the last few seconds, watching the lunar landscape come rushing up at them, they realized their mistake. The camera came down so hard that it smashed the lens against the plaster of Paris. To the astronauts, the impending crash looked all too real.

Sometimes the sims proceeded without the astronauts in the trainer, and one of the people in the S.C.A. had to serve as an AstroSim. The AstroSim was, by the rules of the game, supposed to be able to do anything a real astronaut could do. His job was not a sought-after one. “An astronaut’s got literally hundreds of switches,” Bob Holkan observed, “and he’s got valves he throws, and he’s got sixty pounds of flight-data file that he responds to. This guy [the AstroSim] is sitting there, he’s got a C.R.T., and the ground controllers are asking him, ‘Would you reconfigure the switch on panel A4? And could you tell me the status of the little talk-back over on the left side, and while you’re there, throw the valve.’” That’s the way they tried to do it during a flight—wait until they had several items to package, so they didn’t talk to the crew too much. “Well, that’s nice if you’re in the spacecraft and you’re floating around doing it,” Holkan continued. “But the AstroSim’s got this one C.R.T., and he doesn’t know where in the vehicle this stuff really is. And he’s asking all these other guys in the sim room, ‘What do I tell them? What do I tell them?’”

In these situations, the people behind the picture window were not above lying—“Well, just tell them the switch is already in the right position,” they’d advise the AstroSim. What could the poor flight controller know? “There’s always the possibility that it was a transducer problem, or telemetry problem,” a FIDO pointed out. “The system had enough nodes in it that there was always a possibility that the crew could be telling you one thing and your data could be saying something else and there was a logical explanation.”

Therein lay the continuing, inevitable advantage of the simulation people over the controllers. Under the tough flight directors—Kranz was especially strict on this point—the controllers had to assume that the problem was real. Once in a while, however, an AstroSim would carry things too far. There was, for example, the AstroSim who, during a Gemini sim, was asked by Guido to look out his window and report the position of a particular star. The AstroSim, with no idea where that star was supposed to be, claimed that he was unable to comply with Guido’s request because of cloud cover. That one didn’t wash. But short of that, the AstroSims could get away with a lot. “Play it as it happens,” Kranz preached. “Never accept the fact that it’s a sim screwup.” And so the flight controller would call down to the back rooms and get readouts on obscure possibilities. Finally, when the controllers had thrashed around long enough and had proved to their satisfaction and Flight’s that the situation in the spacecraft couldn’t possibly look the way the AstroSim had told them it did, the AstroSim would say, “Oh yeah, I just found out I hit this other switch with my elbow when I was over here getting out my lunch.” The sims had to be survivable, but they didn’t have to be fair.

Each simulation was followed by a debriefing, “the wake.” The simulation team remained behind their picture window while Flight went to each of the controllers and asked him to describe what had led him to make the calls he had made, and whether in retrospect his calls were the best ones. It was a group confessional, with one controller telling his story, another chiming in to explain how his remarks had influenced the situation, the people in the back rooms describing through the intercom what had been coming up on their screens—“cleansing their souls,” they called it. The debriefings could be as harrowing as the flight, especially when they still had the remote sites. “If you were sitting down in Australia and you screwed up,” Ed Fendell recalled, “Mr. Kraft, Mr. Kranz, or Mr. Hodge would get on the line and commence to tell you how stupid you were, and you knew that every switching center, every office at Goddard, ships at sea, everybody and his mother, everybody in the world was listening. And you sat there and took it. There was no mercy in those days.”

Day after day they went on, one after another, eight or ten launches in a day, eight or ten lunar descents, each with hideous, multiple failures, each with its debriefing and one’s shortcomings laid out for inspection. Bill Tindall was not a flight controller, but he couldn’t stay away. “I used to plug in and watch them go,” he recalled. “It was like going to an exciting movie, watching those people work. Talk about exhausting! Handling emergencies in the missions was an absolute piece of cake by comparison. The real missions were easy.”

Chapter 22. “You’ve lost the engines?”

The morning of April 4, 1968, was one more day in a year that was turning the United States inside out. The Tet Offensive, which had so intensified the debate over the Vietnam War, had occurred two months before. Six weeks before, Minnesota Senator Eugene McCarthy, running for President on an antiwar platform, had shown surprising strength against Lyndon Johnson in the New Hampshire primary. Just the preceding Sunday, on March 31, President Johnson had announced that he would not seek reelection, and in the days since then the newspapers had been filled with speculation about whether New York Senator Robert Kennedy would enter the race. Also in the news, though less prominently, was Dr. Martin Luther King, Jr.’s current effort on behalf of the striking garbage workers in Memphis, Tennessee. This Thursday morning, King was in Memphis, awakening to the new day in his room at the Lorraine Motel.

Within the world of Apollo, the outside world looked completely different. “I missed the entire Vietnam War,” said one, typical of many. “I watched no television, read no newspapers, came to work at six in the morning and worked until nightfall, six or seven days a week for years.”

The people of Apollo were barely aware that the Vietnam War was going on, barely aware that this was a Presidential election year, barely aware that there was such a thing as a War on Poverty or L.S.D. or Sgt. Pepper or race riots. Had one of Chris Kraft’s flight controllers been asked about the most important event so far of 1968, he would probably have said that it had occurred on January 22, when the unmanned Apollo 5 had carried a lunar module on its first test flight. Today, April 4, was going to be the next important date of 1968: the unmanned flight of Apollo 6, the second flight of a fully operational Saturn V.

That morning, Bob Wolf was Booster 1. The Boosters, whose job lasted only the eleven-and-a-half minutes of powered flight it took to reach orbit on a Saturn V (plus another six minutes for translunar injection, for the lunar missions), were always in an odd position in the MOCR. Just when the rest of the flight control team was getting down to work, the Boosters went home.