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

by Kelly, Thomas J.

  To the uninformed observer, however, we looked like outcasts confined to a hidden backwater where we could do no harm to the company’s ongoing business. We worked in a segregated area suspended from the ceiling over a portion of the Experimental Shop, reached by a nondescript flight of dark blue painted metal stairs leading up from the polished wood blocks of the shop floor or by a flight of metal stairs and a catwalk down from an unlabeled door on the second floor Engineering Department office. Both stairways led to a blue metal door in a whitewashed cinder-block wall to which a doorbell and buzzer provided access. Inside, the low-ceilinged compound bathed in fluorescent light and humming with air conditioning seemed like a time tunnel, remote from worldly existence.

  Our office area was cinder-block painted a faded light yellow and crammed with as many wooden desks and chairs as would fit. A single small office at the front of the room, framed by a large window partition, belonged to Al Murder, one of the few permanent members of Preliminary Design. Al was an experienced aircraft designer who had helped fashion Grumman’s Wildcats and Hellcats during World War II and a firm proponent of advancing the company into the space age.

  The group secretary’s voice crackled over the intercom, summoning me to a meeting in Joe Gavin’s office. I put aside the study report I was reviewing and bounded up the stairs to the second floor. It was a welcome change to be on the Engineering floor. The interior vistas were broader and the ceilings higher and neatly finished with white acoustic tiles and frosted glass panes concealing fluorescent light bulbs. Joe Gavin’s office was spacious by Grumman standards. The walls were tastefully covered in rich dark paneling, and it was furnished with dark mahogany furniture with brass hardware and trim.

  Joe and Al Munier were already seated at the conference table, and I joined them there, followed by my deputy Erick Stern. “The LM RFP has been released,” said Joe in his crisp voice with a hint of New England twang. “Saul Ferdman just picked it up in Houston and we’ll have it here in the morning. The proposal is limited to one hundred pages and it’s due in sixty days. Al and I thought we should get together and make sure you have everything you’ll need.”

  Joe Gavin, in his early forties, was a rising star at Grumman. Joining the company in 1946, with an aeronautical engineering degree from MIT and wartime service in the navy’s Bureau of Aeronautics, he soon established a reputation as a talented aircraft designer with leadership capability. As project engineer on the swept-wing F9F-6 Cougar, he directed the design of an improved stabilizer control actuator driven by a high-speed, irreversible, ballbearing screw jack. This novel design allowed the Cougar to safely fly through the Mach 1.0 speed-of-sound barrier in a dive. The Cougar and its straight-wing predecessor, the F9F-5 Panther, both served in the Korean War. Joe became project engineer on the F11F Tiger, the first Grumman-produced fighter that was supersonic in level flight. The Tiger reached limited production and contained many technological innovations.

  Joe Gavin grew up in eastern Massachusetts near Boston. His father was a tinkerer and tool collector, and young Joe developed a keen curiosity about how things work. A visit as an eight year old to a local dirt airstrip where he saw the transatlantic hero Charles Lindbergh impressed him early with the romance of aviation, and his studies at MIT informed him of its technical elegance. By nature reserved and understated, he thought problems through and stuck with his conclusions. Tall, muscular, and with craggy good looks, he was an accomplished oarsman at MIT and an expert skier from his youth. He was a natural leader who, in the face of crises and confusion, remained calm and steadfast of purpose, inspiring others to rally around him.

  Gavin was chief Missiles and Space engineer, in charge of the LM project, and my boss. Al Munier provided the proposal team with support and guidance from the resident Preliminary Design Group. We went over a list of additional engineers we would need and added office space and equipment. High on our equipment list were IBM Selectric typewriters—the new design with the removable type ball—which were in short supply at Grumman. Most of the engineers we needed had worked on our proposal for the Apollo spacecraft in 1961 and on LM studies since then.

  Al, Erick, and I assembled the dozen or so LM proposal engineers in our work area and passed the word. Al promised to find space and desks for another two dozen people in Preliminary Design; the remainder would have to work at their “home” desks and visit us as required. A large conference room in the mezzanine would be available for our daily proposal meetings, which would begin as soon as we had the request for proposal.

  The next morning Joe, Al, Erick, and I joined a standing-room-only crowd in the plainly furnished conference area within Preliminary Design. Saul Ferdman, Grumman’s space marketing director, passed out copies of the RFP and summarized what it contained, having read it on the plane from Houston. It was a drastic departure from NASA’s usual RFP, which normally provided a detailed set of mission plans, spacecraft specifications, and technical requirements and requested that contractors respond with their preliminary design of spacecraft and systems to meet the requirements, their plans for building and supporting the spacecraft to a NASA-specified schedule, and their bid price.

  For the lunar module NASA considered both the mission planning and the technical requirements too uncertain to buy a proposed design. Instead they decided to base contractor selection on an evaluation of which company’s design team was most knowledgeable about the LM’s mission and requirements and had plausible approaches to its design. The company’s manufacturing capability, financial stability, and record of quality would also be considered, and the estimated cost for the program was requested as a means of determining the contractor’s understanding of the program’s scope.

  The RFP was more like a graduate examination in an aerospace engineering design course than a typical government procurement specification. It posed fourteen technical questions and required discussion of five management areas, to be answered in one hundred pages of carefully specified format, even to the type size and line spacing. The technical questions “probed the most exacting technical requirements in the LM mission,” as we told NASA in our response.1 Summarizing some of them:

  1. Discuss the flight mechanics and other considerations of near-Moon trajectories and of lunar launch and rendezvous.

  2. Describe your approach to the design of the following LM systems: onboard checkout, propulsion, reaction control, flight control.

  3. To what extent do you consider backup methods of control and guidance necessary? Describe your approach to this issue.

  4. How do visibility requirements affect LM operations and design?

  5. How would you accommodate micrometeoroid and radiation hazards in the LM design?

  The RFP encouraged contractors to submit a conceptual design of a lunar module with their proposals, as a means of focusing their answers to the questions and demonstrating their competence in manned spacecraft design. However, NASA was not buying the contractor’s design; after the winner of the competition was selected, NASA and the contractor’s engineers together would develop the preliminary design of the LM.

  We already had three conceptual designs prepared against our own estimates of the mission requirements and the space environment. Our major tasks were to compare NASA’s official requirements with our estimates, to determine the impact of differences on our designs, to select a leading candidate design, and to refine and improve that design until time ran out for the proposal. In parallel we drafted answers to NASA’s questions and analyzed them for possible effects on our conceptual design. Within a couple of days we were deeply immersed in this process.

  We ultimately submitted a design for a two-part spacecraft, with a lower landing, or descent, stage and an upper liftoff, or ascent, stage. The descent stage contained mainly the tanks, rocket engine, and plumbing of the descent propulsion system, which was used to drop the LM out of lunar orbit and land it gently on the Moon’s surface; other consumables, such as oxygen, water, and batteries, which cou
ld be left behind on the Moon; scientific equipment to be deployed by the astronauts during exploration; and the landing gear.

  The ascent stage contained the crew compartment and cockpit in which the two astronauts who flew the spacecraft between the Moon’s surface and lunar orbit lived, ate, and slept while on the Moon. This stage contained most of the electronics systems: GNC; communications, radar, and instrumentation; the reaction control system (RCS), with its sixteen small rocket engines that allowed the pilots to control and maneuver the LM in flight; and the environmental control system (ECS), which provided conditioned oxygen and water to the crew for life support and the spacesuits and backpacks needed to go outside the LM’s pressurized cabin. The crew compartment had two access hatches: one on top for docking with, and crew transfer into, the command module, and one forward for crew egress to the lunar surface via a platform. An external docking module allowed either hatch to dock with the CM as a redundancy provision.

  The crew compartment was modeled after that of helicopters, with the astronauts seated in a large forward-facing glass bubble with an instrument panel between them. The landing gear had five fixed legs that just fit within the spacecraft/LM adapter (SLA), inside which the LM was housed at launch from Cape Canaveral. This landing-gear design was what yacht racers would call a “rule beater”: it barely satisfied the critical tip-over and surface-penetration requirements of the RFP without the added complexity of an extendable landing-gear mechanism, its five legs providing enough tread width and pad area within the SLA space envelope to make this possible. Even as we submitted the proposal, I did not expect the real LM to have a fixed landing gear because the slightest change in design assumptions or LM weight would negate this approach, but it was simple and lightweight for the proposal.

  The lunar module proposal design. (Courtesy Northrop/Grumman Corporation) (Illustration credit 3.1)

  The conceptual design met all of NASA’s performance requirements and weighed only twenty-two thousand pounds fully loaded, well below the RFP limit of twenty-six thousand pounds. Grumman’s model shop, accustomed to making beautifully lacquered and detailed display models of airplanes, did the best they could with our ugly duckling, but it remained an alien machine suited to other worlds. The nickname “Bug,” which NASA often used in their studies, still seemed to fit our creation.

  The last of NASA’s technical questions was a zinger: What are the five most important considerations in the design of the LM? List in order of descending importance and explain your reasons for selection.

  I sensed this was a make-or-break question and kept revising our answer until the end. With my technical group I drew up lists of candidate issues and the reasons they were important. Stern, Watson, Gardiner, and I reviewed and debated these and sought opinions widely, from Gavin, Murder, and Ferdman, and from key players, such as Bob Mullaney, Bill Rathke, Bob Carbee, and Arnold Whitaker, who were named in the proposed project organization to join us if we won. At the printer’s deadline, Stern and I finalized the list:

  1. Propulsion design and development

  2. Flight control system design and development

  3. Reliability

  4. Weight control

  5. LM configuration2

  We hoped this would agree with the list Max Faget and his designers had in their minds. Recalling our first meeting on LM with Faget and his group at Langley Research Center, I felt confident that our similar engineering approach and reasoning would lead us to the same answer.

  In the program management section of the proposal, NASA asked for the company’s related experience and performance, our proposed LM program organization and its relationships to the rest of Grumman, our facility and manpower capability, a make or buy plan, and cost estimates. Gavin and Ferdman ran this part of the proposal, but my engineers and I were heavily involved, as our requirements and plans affected everyone else. For example, one of the technical questions asked us to explain our test and development program plans. These established a far-flung series of LM test-beds, test facilities, and qualification programs that consumed a major portion of the time, money, and manpower needed to complete the LMs. Engineers provided the manpower estimates for the design and test phases of the program and part of the manufacturing phase. Making the cost numbers add up and agree with our descriptions of program content required many iterations.

  An aerospace proposal team works as hard as they can to produce the best possible proposal until time runs out. For two months our team of about eighty engineers discussed, debated, analyzed, and wrote our responses to NASA’s exam, cramming as much as we could into one hundred pages of print and illustrations. We included a large foldout general-arrangement drawing of our LM design that showed many of its components, complex functional block diagrams of systems, and a large flow diagram of the development plan. We selected subcontractors and suppliers for LM’s subsystems and major components after holding minicompetitions, and their key people helped with our proposal. The whole team routinely worked fourteen to sixteen hours a day, seven days a week, relentlessly driven by the belief that our competitors were working at least as long and hard as we were and our only chance of winning was to give it everything we had. Like an eight-oared crew, we did not want to have anything left when we crossed the finish line.

  Inside our windowless warren, we lost track of time. One evening when our full crew was hard at work, three maintenance men intruded into our space carrying large stepladders, pipe wrenches, and other tools. Without a word to any of us, they set up the ladders and started dismantling the sprinkler system over our heads. When flakes of rust and then a thin stream of water showered down on the pages I was reviewing, I sprang up to protest. The workmen retreated in the face of my determined objections, but not before telling me who their boss was and assuring me that, whether or not I liked it, the sprinkler system work had to be done.

  The next morning I complained to Al Munier and asked him to get the sprinkler work stopped until work on the proposal had been completed. Returning to the day’s activities, I forgot all about the matter. Late that afternoon, however, Al called me into his office. He upbraided me for being rude to the workmen and said that if I worked more efficiently I would not have to be in the office in the middle of the night. I started to object but quickly stopped. Looking at Al’s face I saw the warning signs of an impending temper tantrum.

  When Al’s temper took hold, his face flushed, his eyes narrowed, and his lips compressed into a hard, thin line. He accused me of arrogance, insensitivity, and inefficiency and told me the sprinkler system had to go in, whether or not I liked it, and while the workmen were there they would clean the light fixtures and paint the ceiling too. Then he abruptly ordered me out of his office.

  Every evening for the next three weeks we faced a daunting succession of distractions and interference. First the plumbers installed new overhead sprinkler pipes, then the cleaners and electricians dusted and wiped the overhead fluorescent light fixtures, adding new fixtures and replacing burned-out bulbs. Then the painters, with their long rollers, white caps, and paint-spattered coveralls, worked on the walls and ceilings. All this sent debris showering down on the desks and floors of our area. It was comical, considering all the work we had to do and our need to concentrate on imagining the design of a unique spacecraft. We survived by packing up our papers and drawings when they appeared, spreading drop cloths over our clean desks, and borrowing desks in an undisturbed area on the far side of the mezzanine. In the brief moments I had time to think, I wondered whether the company understood the importance of winning this contract.

  Two days before the submittal deadline, we turned the proposal over to the document production process. Erick and I gave the galley proofs and graphics a final end-to-end read-through and handed it all to the proposal editor, who took it to the nearby outside print shop for final makeup and printing. Then we joined the rest of the proposal team for a celebratory lunch. I had reluctantly agreed to the team’s request to schedule
the lunch before the proposal was actually delivered to NASA because the Labor Day weekend was approaching and many of our people wanted to stretch it by taking off the preceding Friday. After all the long hours they had put in, I did not want to appear ungrateful. But when I walked into the restaurant and saw our people enjoying cocktails, my stomach knotted up. What if we all suddenly had to go back to work? I never liked having alcohol anywhere near the job.

  I had been in the restaurant about twenty minutes when a waiter told me I had a phone call. It was Saul at the printers. There were two problems: first, when the final page layouts were completed, the book was almost a whole page too long. Something had to be cut. And second, as Saul was reading through the proofs, he discovered an inconsistency between our answers to two of the questions. Which one was correct?

  I thought Saul and I could fix the problems ourselves, and so I quickly headed for the door, but Erick saw me leaving and followed. The printer was in a squat gray building just outside the fence from our area of the Grumman complex. We found Saul with our editor and the printer’s project leader poring over page layouts. Erick and I joined them, and after about an hour we had marked up enough snippets for deletion to bring the proposal within one hundred pages. We were unable to resolve the inconsistency, however, and were forced to call back three experts from the luncheon. We discovered several other statements in our rereading that seemed questionable and had to be verified with the people who wrote them. Before long we had about ten people in the printers, trying mightily to focus on minute details of the proposal after coming from what had developed into a roaring party. To their credit, they were soon able to satisfy me and Erick about their sections or make minor modifications for added clarity, but it was a frantic couple of hours before all the loose ends were secured and the proposal was again declared finished.