Moon Lander: How We Developed the Apollo Lunar Module (37 page)

LM-1 was put into a perfect orbit by its Saturn 1B booster, Apollo-Saturn 204, the same rocket that had been assigned to launch the ill-fated Apollo 1. The spacecraft/LM adapter and the nose cone were jettisoned as planned, and the unmanned LM’s systems were activated by ground commands from Mission Control in Houston. After checking out the spacecraft for two orbital revolutions, Mission Control commanded the first major systems demonstration of the flight; a thirty-eight-second firing of the descent engine. However, after four seconds LM-1’s guidance system shut down the engine because it had not provided sufficient acceleration to satisfy a cutoff threshold built into the software. Consulting with my guidance experts at KSC, Houston, and Bethpage, we quickly established that this was simply a software error that did not indicate any problem with the descent propulsion system. In a normal descent propulsion start-up the propellant tanks would be fully pressurized before the firing command was initiated, and the acceleration would exceed the required threshold in less than four seconds. However, LM-1’s first descent propulsion firing began with the tanks only partially pressurized as they were at Earth launch, so it took an additional two seconds to reach full tank pressure, engine thrust and spacecraft acceleration. The software cutoff should have been changed to six seconds for the first firing, but it was not, causing the unwarranted engine shut down.

NASA Houston agreed with our conclusion that there was nothing wrong with the LM, and switched to a backup plan using ground commands to the LM mission programmer to salvage the mission. Continuing with the preset
in-flight commands that had been programmed into the LM guidance computer’s software was no longer feasible because the timing of the mission events had been upset by the need to repeat the aborted descent engine firing. Event timing was critical in Earth orbit because the spacecraft must be in contact with a ground station of the worldwide tracking network whenever flight activities occurred. The LMP, a special piece of equipment for unmanned LMs only, allowed LM systems operations and maneuvers to be commanded from the ground in any desired timing and sequence.

Transitioning smoothly into this well-rehearsed fallback mode of operation, Flight Director Gene Kranz’s firm, confident voice over the net led the rearrangement of mission events to assure that, if the LM were capable, all flight objectives would be met. Kranz and his flight controllers consulted frequently with the Grumman support staff, and we all concurred on their revised flight plan. Houston executed the plan crisply and successfully performed all mission events, including prolonged descent and ascent engine burns, and FITH ascent engine start-up and separation from the descent stage. After cross-checking that all mission objectives had been accomplished, LM-1’s ascent stage was shut down for reentry less than eight hours after liftoff from KSC. It was a brilliant recovery from an uncertain beginning.
²

I was elated by LM’s basically good performance in its first flight in space. Witnessing a launch brought home to me the full complexity of Apollo missions, and how totally dependant they were on every one of the thousands of people involved doing his or her job right. Trust in the quality of the work performed extended far upstream of the final assembly process and down to the final launch preparations at KSC. Apollo’s successes would be the proud result of the individual efforts of thousands of engineers, managers, crafts people, and artisans, but a critical moment of carelessness or inattention by any one person could cause failure. It was a sobering thought as we prepared to move into manned LM flights.

14

The Dress Rehearsals

Apollos 9 and 10

The fateful year arrived—1969, the final year of the decade chosen by President Kennedy for America to make history in space and show its technological supremacy over the Soviet Union. After missing Apollo 8 the LM still had not been flown by astronauts in space, but NASA plunged ahead bullishly, counting on success, and scheduled launches every two months beginning in March. Three flight LMs (LM-3, LM-4, and LM-5) were at KSC undergoing checkout and prelaunch preparations. The high-bay spacecraft assembly clean room in the O&C building pulsed with around-the-clock activity: three command modules and service modules were on the floor also, tended by their own legions of engineers and technicians from North American.

First Manned LM Flight: Apollo 9

The tardy LM was the program’s big unknown. Data from the unmanned flight looked good, but LM lacked the assurance that comes from having sharp-eyed astronauts living aboard in space, flying, probing, noticing every detail of its in-flight performance, up close and personal. Wally Schirra and Frank Borman and their crews made many observations and suggestions in their debriefings from the Apollo 7 and 8 missions that helped NASA and North American make small improvements to the spacecraft, equipment stowage, and mission operations techniques that bolstered confidence in the CSM’s readiness to head for the Moon. LM needed this also, and it had to show that it could perform the critical mission functions under pilot commands.

Building and Testing

LM-3 was designated to be the first manned LM, and was first to receive all the materials changeouts, quality procedure enhancements, and design changes
resulting from the Apollo 1 fire. This slowed its progress in Spacecraft Assembly and Test in Bethpage, but by fall of 1967 the basic manufacturing and assembly was completed and the ascent and descent stages were mated together in the combined stages workstand. Further equipment and component installations continued, and the spacecraft was connected, by white “interface boxes” on the assembly floor and workstand platform levels and many thick, sinuous black cable bundles, to the automatic checkout equipment station that overlooked the floor of the clean room. This allowed formal acceptance testing of LM-3 to begin even as further manufacturing assembly continued.

As LM-3 moved further into the checkout sequence, the tests required interaction with the pilots in the LM cockpit. Grumman test engineers and LM project pilots Jack Stephenson and Scott MacLeod fulfilled this function, which could be required at any time during S/CAT’s three-shift, twenty-four-hour workday. NASA’s flight crews intended to participate in these tests, and we set up trailers in the parking lot behind Plant 5, adjacent to the high-bay spacecraft Assembly and Test clean room, as their office and motel. For each manned Apollo flight, NASA assigned three crews: the prime crew, who would fly the mission unless ill or incapacitated before launch; the backup crew, who would assist the prime crew in preparing for the mission, representing them at meetings, briefings, inspections and tests, and substituting for a prime crew member on the flight if required; and the support crew, who represented both the prime and backup crews as needed to cover the many simultaneous program events and activities in which the crews were interested. For Apollo 9, the crew assignments were as follows:

We saw mainly McDivitt, Schweickart, Conrad, and Bean on LM-3. The flight crews spent more and more time at Bethpage, sometimes standing for hours in the LM crew compartment as an OCP slogged through its lines. They delved into their LM with intensity: questioning everything, reviewing equipment test records from the suppliers, visiting key subcontractors, and acting as the scripted pilots in major tests. Their curiosity, persistence, and endurance knew no bounds.
¹
Stephenson and MacLeod helped them at Bethpage, performing tests for them in the LM cockpit and following up to obtain answers to their lists of questions.

Jim McDivitt was a good-looking fellow, above average in height and with a trim build. He had a pleasant face, a ready smile, and quizzical eyebrows that could point upward toward the center of his forehead in silent incredulity when offered an unsatisfactory explanation. He was a veteran astronaut, having
been commander of Gemini 4, a trailblazing flight that included a rendezvous experiment and a highly successful rendezvous exercise.
²
He had also been the backup commander for the ill-fated Apollo 1. Jim had a wide understanding of aerospace systems and design and argued tenaciously with me and my engineers about any aspect of LM that did not seem right. He roamed the assembly clean room and shop floors widely when at Bethpage, poking into obscure corners and questioning the workers about what they were doing and why. Although unfailingly polite, Jim would not tolerate excuses or evasions.

“Rusty” Schweickart was a rangy, rawboned young man with short, reddish hair, a ruddy, freckled complexion, and the open, earnest expression of one eager to learn and perform. Rusty wanted to know absolutely everything about his LM, and he patiently studied system diagrams, operating manuals and test procedures until he understood how each system and component worked. Then he could recognize normal system performance and visualize operations in backup and degraded failure modes. As the LM pilot he spent many hours in the LM mission simulators at Houston and at Kennedy Space Center, learning to fly the LM in all mission phases and failure conditions. McDivitt had to fly both the lunar module and the command module simulators, but Schweickart could concentrate on the LM, becoming an expert pilot, able to handle any emergency.

Simulators and Combat Boots

The mission simulators for the LM and the CM were workhorses of the astronaut training program. Both were made by Singer-Link, successor to Link Aviation, which made the famous Link Trainer, the rudimentary cockpit flight simulator that was the first step toward the skies for thousands of World War II pilot trainees. We placed the LM simulator under contract a year and a half after Grumman’s go-ahead, once the preliminary design of the spacecraft firmed up. It was a complex computer-controlled flight simulator with accurate replication of the LM flight station, controls, and displays. Realistic lunar surface scenes were projected onto the windows by an optical system developed by Farrand Optical Company. This innovative system used a small fiberoptic camera that “flew” over a three-dimensional plaster model of the lunar landing site (or the Earth for Apollo 9), remotely driven by computer commands that matched the LM’s flight path.

The LM mission simulator could insert faults into the LM’s systems, displaying the resultant instrument readouts, caution and warning alarms, and effects on LM’s operation and flight performance. Bogus failures were entered into the LMS from an instructor’s console outside the simulated crew compartment. It was a fixed-base simulator—there was no motion of the crew compartment, but an electrical vibrator mimicked rocket firing and audio
tapes played cabin noises, such as environmental control system fan and pump hums and reaction control system thruster firings. The LMS could replicate LM’s communication systems accurately, providing blackouts from contact with Earth when LM was behind the Moon and static and dropouts if LM’s steerable antenna lost lock or the omni antennas were in unfavorable positions. There were two LM mission simulators, at MSC Houston and KSC Florida, and they were booked weeks ahead for flight crew training and mission simulation support.

Grumman also designed and developed the full mission engineering simulator (FMES), an in-house Grumman guidance and control system simulator used for engineering development and integration of the LM’s guidance, flight controls, and computers. The FMES had a rudimentary flight station that did not stress realism and lacked optical displays, but it had a highly accurate computer-controlled flight attitude table on which inertial measurement components (platforms, gyros, accelerometers) were mounted and subjected to flightlike environments. Prototype flight hardware was used in these tests, and during flight missions the FMES proved a vital problem-solving resource.

One Sunday winter morning McDivitt came charging into my small office in a trailer behind the LM Assembly Hangar in Bethpage and shouted, “Hey, Kelly! Do you know there are guys wearing combat boots in the LM cabin? They could put their feet right through the cabin skin if they’re not careful.”

I followed him into the LM assembly hangar. Everyone entering had to don white smocks, caps, gloves, and booties, but we allowed them to put the cloth booties on over their shoes (not expecting boots!). McDivitt was right; I found at least three boot wearers myself. We changed our factory dress rules to require that shoes be left in the entry area and the booties put on over socks.

Passing Muster at KSC

After passing an intense two-phase customer acceptance readiness review at Bethpage, LM-3 was shipped to Kennedy Space Center on 14 June 1968—my birthday, which I took to be a good omen. To prevent a repeat of the LM-1 receiving inspection debacle at KSC, we arranged to have a team of NASA and Grumman quality inspectors work at Bethpage for several weeks before delivery, sharing inspection duties with the resident team in S/CAT. Both inspection teams signed off on the delivery acceptance papers. (LM-2, not required for flight because of LM-1’s success, was delivered to NASA for storage and ultimately displayed in the Smithsonian National Air and Space Museum in Washington, D.C.)

Despite this precaution more than one hundred discrepancies were found in LM-3 during receiving inspection at KSC. Broken wires and structural cracks due to stress corrosion were of the greatest concern. LM-3 contained
thin, 26-gauge wires with low-strength annealed copper alloy, which were very delicate. The failure of a window on LM-5 during a pressure test at Bethpage cast further doubt over LM-3, prompting intense inspection of its windows while the engineering investigation proceeded at Corning Glass.