Moon Lander: How We Developed the Apollo Lunar Module (Smithsonian History of Aviation and Spaceflight)

by Kelly, Thomas J.

  At this stage of the Apollo program, much more than piloting skills were required of the astronauts. They also had to be competent field geologists, capable of representing the Earth-bound scientists by conducting efficient, perceptive surface traverses in limited time in a hostile environment. Although temperamentally the opposite of the patient, painstaking lunar scientists, Shepard resolved to attain the knowledge and skills needed to be a discerning field geologist. Leading his crew by example, he plunged into field training in the mountains of California and Arizona under Lee Silver, who together with geologist-astronaut Harrison “Jack” Schmitt was giving the Apollo astronauts a crash course in field geology. They learned to identify minerals and to describe what they saw in precise geological terms. They also studied the different theories on the origins of the Moon and how the specific objectives and possible findings of their mission might confirm or refute them. As in all else he did, Al Shepard was resolved to excel in this new dimension of an astronaut’s job.4

  Shepard hand-picked his crew, drawing upon the insight gained into their capabilities from his role as chief astronaut. He selected two rookies who had neither flown in space nor served on a backup crew. He delegated to them the responsibility of becoming experts in the systems aboard their spacecraft, leaving himself free to concentrate on flying and mission operations.

  No longer the stern “Big Al” who had lorded it over the others as chief astronaut, Shepard the mission commander was down in the trenches with his crew, soaking up simulator time, conducting ground tests on their command module and lunar module, and clambering up cliffs in geology encampments. His crew mates found him pleasant to work with, even considerate, but still thought of him as a professional associate rather than a pal.

  Edgar Mitchell, LM pilot, had a doctorate in aeronautical engineering from MIT and had been an instructor at the Air Force Test Pilot School at Edwards Air Force Base. He was thoughtful and intellectual, soft-spoken, but could become impatient and exhibit a hard edge of anger. He became expert on the LM’s systems and their possible failure modes, and I often saw him in Bethpage, roaming the floor in Spacecraft Assembly and Test, examining flight hardware and discussing its idiosyncracies with the technicians who installed and operated it. Mitchell conducted many tests at Bethpage in LM-8’s cockpit and was quick to call to my attention any deficiencies he encountered, whether in equipment performance, test procedures, or the Grumman test personnel. His complaints were constructive, involving specific, real problems that we could correct.

  CM pilot Stuart Roosa had been an air force pilot in Germany preparing for nuclear war, and later a test pilot at Edwards. Eager and competent, he learned the Saturn booster’s systems before his assignment to Shepard’s crew, working with von Braun’s engineers at Marshall Space Flight Center in Huntsville, Alabama. As CM pilot, he delved into the command and service modules’ systems with even greater intensity. He appreciated the free rein that Commander Shepard gave him to spend time at North American’s factory in Downey, California, learning the intricacies of his spacecraft.

  Shepard’s team trained together for nineteen months, longer than any other Apollo crew, due to their reassignment to Apollo 14 and the program delays following the aborted Apollo 13 mission. They enjoyed an easy and correct professional relationship but were not the close buddies that the members of Conrad’s and Lovell’s flight crews had been. They confirmed Deke Slayton’s theory that he could assign any three astronauts together to form an Apollo flight crew, no matter how diverse their personalities, because they were all superb professional pilots.

  On 31 January 1971 Al Shepard and his crew, LM pilot Edgar Mitchell and CM pilot Stuart Roosa, were strapped inside the command module Yankee Clipper ready to blast off to the Moon. At forty-seven Shepard was determined to prove that he was the world’s best test pilot and astronaut. His Apollo 14 was almost fully dedicated to science and exploration, and Shepard and his crew were committed to every one of its long list of objectives.5

  After being successfully boosted into Earth orbit by the Saturn 5, and injected into the translunar trajectory by the Saturn’s S-4B stage, Apollo 14 encountered a problem that could have ended the mission before it had fully begun. Kitty Hawk was unable to attain hard docking with LM Antares, despite four attempts over a period of one and a half hours.

  From my post in the Spacecraft Analysis Room in Mission Control, I discussed the problem with our mechanical systems section leaders, Jiggs Sturiale and Marcy Romanelli. They examined inspection records and photographs of LM-8’s docking mechanism and docking ring structure, and had many discussions with their NASA and North American counterparts. Jiggs and Marcy considered debris in the mechanism, or adverse dimensional tolerance buildups the most likely causes of the problem.

  Some way had to be found to apply more force to the docking assembly to compress it slightly, allowing the latches on the CM’s docking ring to snap home, or the mission would have to be aborted. A solution was required in a few hours, before pressure buildup in the S-4B’s tanks would cause them to vent, sending the S-4B, and the attached LM, spinning out of control.

  The docking system engineers came up with a modified procedure for Shepard to try. Normal docking procedure was to engage the probe, mounted in the CM’s docking tunnel, with the drogue inside the LM’s tunnel. Spring-loaded latches on the probe locked into slots inside the mating hole in the drogue, holding the CM and LM together in a “soft-docked” condition. Then an electric motor drive retracted the probe several inches until the spacecrafts’ two thirty-two-inch-diameter docking rings made firm contact, automatically activating the CM’s twelve docking ring latches to clamp tightly against machined pads on the back side of the LM’s docking ring. This produced a “hard dock” in which sizable mechanical forces and moments could be transmitted across the joint between the CM and the LM, allowing the two spacecraft to be maneuvered together under control of the thrusters in the CSM. Kitty Hawk and Antares had achieved soft dock, but they were unable to obtain hard dock.

  The engineers’ solution was to fire the CSM’s thrusters to force the two docking rings tightly together, and then retract the probe while the thrusters continued to fire. The additional compression force applied by the thrusters might overcome the dimensional problem that was keeping the latches from snapping home on the pads.

  Roosa lined up the docking target carefully, then fired the thrusters while separated about two feet from Antares. The crew felt a firm thump as the docking rings mated. Shepard retracted the probe. At first nothing happened, but then his panel display flipped to the striped “barber pole” position indicating latches engaged, and the crew heard the reassuring “ripple bang” sound of twelve latches snapping shut. Apollo 14 was still “go” for the Moon.

  After the mission, Al Shepard admitted that he had considered riskier solutions if the engineers’ fix had not worked. He thought about donning spacesuits, depressurizing the cabin, and removing the probe and drogue from the tunnels for inspection and cleaning. Or, alternatively, using their gloved hands to apply added clamping force to the two docking rings. He was not about to be denied walking on the Moon by such a silly hangup.6 Recalling his delight at Corky Meyer’s quick and dirty test of the overpressure in the Cougar’s nose, I found his fall-back docking fixes right in character.

  The head-to-head docked spacecraft proceeded smoothly on the long coast to the Moon and were injected into lunar orbit by the large service propulsion rocket engine. Shepard and Mitchell entered Antares in their spacesuits and activated her systems. In the SPAN Room I followed the detailed LM flight plan closely, noting the mission time at which planned actions occurred, such as transfer to LM power, S-band steerable antenna activation, landing-gear deployment, and reaction control system pressurization and hot-firing checkout. As they went behind the Moon and out of touch with Houston, the crew had reinstalled and verified the docking probe and drogue in the tunnels and closed the hatches on the two spacecraft.

  A fe
w minutes after emerging from behind the Moon, Antares undocked from Kitty Hawk and revolved gracefully for Roosa’s inspection. They moved apart as Shepard conducted a test firing of the LM descent engine and pitched over to observe and photograph the landing site.

  I was happily checking off LM flight plan events within a few minutes of schedule when I saw something that made my stomach knot up. The LM guidance computer reported receiving a signal to abort the mission. It could only have come from the abort-stage switch on the LM’s control panel. The computer ignored the abort signal because it was programmed only to respond to such a command during powered descent. MIT confirmed that the LGC did receive the errant signal; however, it did not show up on the LM’s instrumentation readouts.

  I conferred in SPAN with Bob Carbee and Arnold Whitaker, and we alerted our guidance experts and our cognizant engineer on the abort-stage switch. After consulting also with NASA and MIT, we recommended a special procedure to attempt to clear the bogus signal before Antares went behind the Moon: Shepard pressed and held the engine-stop switch while pushing and resetting the abort-stage switch. The errant bit in the computer disappeared just before we lost radio contact.

  A whirlwind of activity struck Houston, Bethpage, and Cambridge as NASA Mission Control, Grumman, and MIT debated what to do when Antares emerged again in fifty minutes. In SPAN, Carbee, Whitaker, and I pored over a schematic diagram of the abort-stage switch; in Bethpage two such switches were pulled from the stockroom and installed into test fixtures. At MIT, programmers studied the LGC’s software instructions related to abort. The abort-stage switch was a push button with fifteen electrical contacts in a hermetically sealed case. It was a “panic button” for quickly performing an abort-stage maneuver during LM-powered descent—shutting down the descent engine, separating the ascent and descent stages, starting the ascent engine, and pitching up to an ascending trajectory for rendezvous with the command module. Only two signals from the switch went to the computer: one told the computer to shut down the descent engine and start the ascent engine, the other changed the weight and mass properties in the computer to that of the ascent stage only. These signals were of great concern, because if erroneously present during powered descent, the LGC would act upon them. The other thirteen signals activated the explosive devices for stage separation, but they could be blocked by leaving the explosive devices Master Arm switch in the “off” position.

  Mission control decided to ask the crew to tap the instrument panel near the switch. After radio contact was reestablished, Ed Mitchell began tapping. Our worst fears were confirmed: as we watched the telemetry readout of the fifteen switch signals, one or another of them randomly changed state (closed or opened) as Mitchell tapped the panel. A solder ball or other conductive debris was floating around inside the abort-stage switch, completing the circuit one signal at a time, depending upon where it alighted.7 The only solution was to devise a software change that would tell the LGC to ignore signals from the abort-stage switch. MIT had less than ninety minutes to come up with the fix.

  Within a few minutes a young programmer at MIT named Don Eyles wrote the corrective software and tested it on the LGC in MIT’s laboratory. He transmitted it to NASA and Grumman, where it was independently verified on the LM mission simulator at Houston and the full mission engineering simulator at Bethpage. When Antares reappeared, Mission Control relayed the instructions to Mitchell, who keyed it into the LGC a few minutes before the start of powered descent. Once again the ground-support team had saved the Apollo 14 mission.

  Shepard fired the descent engine and began the eleven-and-a-half-minute powered descent. The engine started and burned smoothly, throttling up and down as directed by computer commands, and Antares’ altitude above the surface rapidly decreased. About six minutes into the descent the LM passed through thirty thousand feet, and the crew looked for landing radar data. It did not appear, even at twenty-five thousand feet, the specification requirement. The landing radar had checked out okay in lunar orbit. What could be the problem?

  Mission rules required an abort if the LM had no landing radar data below ten thousand feet, in about two minutes. I took a deep breath. The LM guidance flight controller recommended opening and closing the landing radar circuit breaker to reactivate the radar’s start-up sequence. Mitchell tried it, and good landing radar data filled Antares’ display and Mission Control’s screens at twenty thousand feet. Carbee, Whitaker, and I cheered and grinned with relief, wiping imaginary sweat from our brows. Another mission save with ground assist!

  Shepard pitched over at eight thousand feet and saw his main landmark, Cone crater, dead ahead where it should be.

  “Fat as a goose!” he told Houston, and steered Antares to a smooth touchdown, closer to the target than any other landing on the Apollo program. It had been a long, uncertain journey, but the undaunted warrior and his crew mate had arrived on the Moon.8

  Apollo 14 carried an expanded version of the Apollo lunar surface experiments package and a hand-drawn two-wheeled tool carrier that Shepard called the “lunar rickshaw” (jargon-loving NASA named it the modular equipment transporter, or MET). In the first moonwalk, the explorers deployed the ALSEP and gathered soil and rock samples from the Fra Mauro plain near Antares. They gazed in wonder at the broken, boulder-strewn gray plain, the black sky, myriad bright stars, and stunning blue-and-white Earth overhead. They looked up at the forbidding steep ashy slope leading to Cone crater’s rim, which would be the main goal of their second traverse. They selected, described, and documented the rocks and other geological features of most interest to the scientists, living up to their promise to excel at all aspects of their mission.

  In SPAN I followed them step-by-step, noting in my copy of the flight plan the actual time at which they performed each lunar surface activity. I heard them report on the Fra Mauro soil, “The surface layer is very soft, fine, and clings to everything. It looks like brown talcum powder.”

  “The rickshaw is leaving tracks about three-quarters of an inch deep. They glint in the sunlight, and we can see them all the way back to the LM,” they said. The astronauts set out an acoustic geophone array and tested it with explosive soil-thumper grenades, leaving a mortar to be fired remotely after they departed. Their precise landing enabled Apollo 14’s crew to perform the traverse route and activities they had discussed with the scientists in Houston weeks before.

  Once inside Antares, Shepard and Mitchell doffed their helmets, gloves, and life-support backpacks, weighed and stowed their samples, ate dinner, and settled down to sleep. Although ten hours of rest was allocated to prepare them for the physically demanding second moonwalk, eight hours was enough for this pair, and they emerged from Antares two hours ahead of time. They had a difficult climb up the steep slopes of Cone crater, sliding on loose rock and ash, with the MET often tipping over downslope. At times they had to carry it between them. More than 1,100 feet in diameter, Cone crater loomed 250 feet above the plain but was not visible from below. Its rim was expected to be covered with rocks and boulders ejected from deep beneath the surface in the earliest days of the Moon’s existence.

  Despite a valiant effort, Shepard and Mitchell did not reach the rim and gaze across the crater’s vastness. The lack of size reference and memorable landmarks, the sharply undulating surface and unfamiliar lighting conditions combined to make the ascent of Cone’s flanks seem like an outdoor hall of mirrors to the frustrated explorers. Mitchell went over to Shepard and showed him the map. “Look,” he said, breathing heavily, “let me show you something.… We’re down here. We’ve got to go there.”

  They reached a level area, but the rim was still nowhere in sight. Mitchell again studied the map. “This big boulder, Al, that stands out bigger than anything else—we oughta be able to see it.”

  “Okay, Ed and Al.” From Houston the firm voice of CapCom Fred Haise told them that time was up. They had used the half-hour extension that Mission Control had granted and still not found Cone Crater. In their st
iff pressurized spacesuits, they were tired from the exertion of climbing uphill. It was later determined that during their return they were within sixty-five feet of Cone’s rim, which looked to them like just another of many false summits.

  Despite not finding Cone’s rim, Apollo 14 was a major scientific success. The explorers broke all prior records for time on the Moon (one day, nine and a half hours), duration of moonwalks (nine hours, twenty minutes), and pounds of samples returned (94.4). They squeezed the most they could from Antares, the last of the basic LM designs to fly in Apollo. Subsequent missions would have the extended stay version of LM, capable of spending three days on the Moon, and carrying the Lunar Roving Vehicle.

  Before reentering Antares, Shepard produced a golf club and balls, and standing in front of LM’s TV camera, he showed the world how far one could hit a golf ball on the Moon, in one-sixth gravity and without atmospheric drag.

  “In my left hand,” he announced, “I have a little white pellet that’s familiar to millions of Americans …” On his third try he hit the ball solidly, lofting it up into the black sky and over the craters in slow motion.

  “Miles and miles and miles!” Shepard chortled.9 Without that clumsy spacesuit he felt he could have put it into orbit.

  While the crew was tidying up Antares’ cabin and eating, in SPAN I was working on a problem in the S-band steerable antenna, the LM’s primary communications link to Earth and the only means of high-bit-rate data transmission. The antenna was jittering around its locked-on position when transmissions were sent or received. We feared that it might break lock during ascent, depriving Mission Control of real-time data. NASA and Grumman engineers stopped the jitter by switching to an alternate ground uplink mode and recycling the antenna mode switch in the “auto” position. In any event, the omnidirectional antennas on LM would provide a voice link and low-bit-rate data.

  Shepard and Mitchell proceeded through the LM prelaunch checkout. When they test-fired the reaction control system thrusters the exhaust breeze knocked over the erectable antenna that had been used to provide TV and high-bit-rate data and voice during the moonwalks. Liftoff took place on schedule, and the seven-minute ascent engine burn was flawless, placing Antares in the planned ascending rendezvous orbit.