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

by Kelly, Thomas J.

  In hindsight the net result of the Apollo 1 accident was an eighteen-month slip in the CSM schedules and four months’ slip in LM. This allowed LM, which had been about one year behind the CSM and Saturn, to catch up. At the time this was not clear to me or anyone else at Grumman, and we fought tenaciously to preserve every day of our schedule despite the many problems and obstacles.

  Reorganization and Recovery

  Major management changes took place in NASA and North American as a result of the fire. George M. Low, formerly deputy director of the Manned Spacecraft Center, accepted a demotion in rank but a promotion in real authority and responsibility to become Shea’s replacement as ASPO manager. He soon instituted the powerful Change Control Board, which had to grant approval before any significant change could be made to the Apollo spacecraft. The CCB, whose members included Kraft, Faget, Bill Lee (later replaced by Bolender),7 Kenneth Kleinknecht, Deke Slayton, Tom Markley, and George Abbey, met weekly to review and approve changes proposed by NASA or the prime contractors. Low made the final decision on each item, in his habitual careful, cautious, well-reasoned fashion, after patiently soliciting and hearing all viewpoints. The CCB met weekly and many meetings lasted all day and well into the night, but with rare exceptions the premeeting agenda was dealt with. The CCB became George Low’s primary tool for controlling the spacecraft program at a detailed level.

  Gilruth assigned astronaut Frank Borman to be NASA’s on-site program recovery team leader at Downey, and Low spent considerable time there himself, helping North American sort out and correct its problems. In helping them solve their problems, Low formed strong sympathetic bonds with North American’s executives and managers that continued beyond the Apollo program.

  At NASA’s insistence, Harrison Storms was removed as president of North American’s Space Division. His replacement was William D. Bergen, an experienced aerospace executive recruited from the Martin Company. Bergen brought with him two associates from Martin: Bastian “Buzz” Hello to run the Florida launch support operation for North American and John P. Healey to manage assembly and test of the first manned Block 2 (LM compatible) command module at Downey.8 The capable veteran Dale D. Myers remained as North American’s Apollo program director.

  Bill Bergen took command at Downey; he evaluated and replaced key personnel and reviewed and overhauled organizations and procedures. One major action was to set up personalized teams to assemble and test each spacecraft in the factory. John Healey, former Manufacturing manager at Martin, was the spacecraft team manager (STM) for CSM-101 and the role model for spacecraft team leadership. Tall, dynamic, and confident, Healey assumed total responsibility for his spacecraft and had the full support of Bergen and Myers. He knew the aerospace manufacturing business and was uncompromising on quality but hard-driving on schedules. Best of all, he got results. Borman and other astronauts were impressed and believed that spacecraft 101 would set a new standard for quality, far above that of its ill-fated predecessor 012. This management approach was soon adopted by Grumman when Evans, Titterton, and Gavin appointed STMs to direct the assembly and test of each lunar module.

  Apollo flights resumed in November 1967 with Apollo 4, the first flight of the gigantic Saturn 5 booster rocket. It launched an unmanned Block 1 Apollo spacecraft, the refurbished CSM 017. The mission successfully demonstrated Saturn 5 performance, except for the Pogo problem, which had been encountered in other launch vehicles and immediately became the focus of engineering attention at Marshall. CSM flight objectives were fully accomplished, including such critical items as demonstration of service propulsion system (SPS) performance and restart capability and verification of CM heat-shield adequacy for lunar return reentry.

  At last it appeared that the Apollo program was on the way to recovery.

  12

  Building What I Designed

  George Skurla, Grumman’s director at Kennedy Space Center, entered Col. Rocco Petrone’s spacious office, his lips pressed tight and eyes narrowed into slits. Skurla knew what was coming. LM-1, the first flight lunar module, had been delivered from Bethpage the day before, 21 June 1967, and had immediately failed propulsion system leak tests during receiving inspection. He would get a blunt “chewing out” in Petrone’s army company commander style.

  Petrone, tan, muscular, and still rock solid like the West Point football player he was, rose from behind his large walnut desk to full imposing height and expanded his broad chest, his face twisting into a dark scowl. As NASA’s director of Launch Operations for the Apollo program, Petrone ruled everything on the cape with an iron hand, especially the contractors who supported his launches. He was skeptical of Grumman, the newcomer to space and to the Apollo program, and intolerant of the fast-talking New Yorkers with their atrocious accents and streetwise, independent ways. And now he saw that they could not deliver a quality spacecraft.

  “George,” Petrone said, “what kind of two-bit garage are you running up in Bethpage? That LM you sent us yesterday is supposed to fly in space, but I wouldn’t even allow it on the launch pad. Its propulsion tanks and plumbing leaked like a sieve—it’s a piece of junk, garbage! You should be ashamed. And four months late besides.”

  Skurla averted his eyes and pressed his lips so tightly together they turned white. A veteran Grumman flight test engineer, he had directed the structural flight test demonstration of many Grumman airplanes, including the F9F Panther and Cougar jet fighters, and the large E2C Hawkeye with its dominant rotating radar antenna. He was intensely proud of Grumman and felt that he personally carried on the tradition of the “ironworks.” LM-1 was an embarrassment, and he felt betrayed by the people at home.

  “Now take it easy, Rocco,” Skurla said. “It has some problems, but it’s not junk. Nothing with the Grumman name on it is junk.”

  “When they turned on those sniffers there were sirens wailing everywhere,” Petrone continued. “What kind of so-called tests did they do in New York before sending this wreck to us? You guys were supposed to be a cut above North American, but now it seems you’re even worse. I want you to tell Evans and Gavin that NASA won’t stand for this. They had better get this fixed, and fast! Your name is mud around here until they do.”

  Skurla retreated to his office and called Gavin. “Joe,” he said, “I just sat through the most uncomfortable meeting of my life. Rocco Petrone chewed me up, down, and sideways, and you know something—he was right. You guys really let me down, every one of you. Especially that guy Kelly—wait ’til I get hold of him. How could he send a spacecraft that leaks like a sieve? ‘Junk, garbage,’ that’s what Petrone called it.”

  Later Skurla was on the phone to me. I had to hold the receiver away from my ear, he was yelling so loudly: “Listen, Kelly, I want you to know you’ve disgraced the Grumman company, and made my life miserable down here besides. Petrone told me you sent us ‘junk, garbage—leaks like a sieve.’ What the hell is the matter with you guys? Don’t be too proud to get help—talk to Radcliffe, he knows all about leaks. Get Corky Meyer and Flight Test to help you. Don’t keep everything locked up tight within Engineering.”

  “I’m sorry, George,” I said. “I’m really sorry. All I can say is it didn’t leak when it left here. We must be doing the leak testing differently—we’ll have to straighten that out.”

  “You’d better straighten everything out, and fast! I want this fixed, and I’ll get Gavin and Titterton and Evans and God Himself if necessary to make sure it happens. So smarten up, kid. Get off your ass and do what it takes to fix those leaks. And one other thing—this had better never, never, ever happen again!”

  Skurla hung up abruptly, leaving me to wallow in gloom and self-recrimination. My S/CAT team and I had let Grumman and Apollo down. We were a laughingstock, the butt of scorn, pariahs of the program. Yet just a few short days ago we had worked our hearts out on LM-1 and were proud of passing NASA’s high hurdles for delivery at the customer acceptance readiness review. What had gone wrong?

  Shakeup
>
  It began with a fierce snowstorm that struck Long land in the darkness on Tuesday morning, 7 February 1967. I could hear the wind howling outside, pummeling our bedroom, which was on the storm-targeted northeast corner of the house. It was snowing hard, with perhaps six inches on the ground already. I dialed the Grumman snow closing number and a smile of relief crossed my face as I heard the recorded voice declare that all Grumman plants were closed due to the storm. Hooray! Instead of battling traffic at 7:00 A.M. I could crawl back into my warm bed and look forward to a day of playing in the snow with Joan and the children. I drifted off to sleep again without even a twinge of conscience.

  I was exhausted from an endless succession of six- and sometimes seven-day weeks, with twelve- to fourteen-hour days. The workload was very heavy now that much of our system and component hardware was being tested at the suppliers and encountering failures, and our first flight LM was in manufacturing and test, along with other flight LMs and test articles. But beyond fatigue was the shock and dejection that our LM team felt when, just ten days previously, we had learned of the tragic fatal fire in CM 012 on Pad 34 at KSC. The whole program was still reeling from its impact, and there was a degree of confusion as NASA upper management wrestled with the implications of this accident and the extent of corrective actions and program rescheduling that would be required. I was relieved to gain an unexpected day off.

  Back at Grumman the next day I was summoned to a special meeting called by George Titterton and Joe Gavin. All the upper management of the LM program was there, and much of Grumman’s top corporate management. Titterton sat alone at the plain plastic-topped metal table in the large Plant 25 conference room in a long-sleeved white shirt and dark tie, staring balefully at the expectant audience through his thick glasses. The crowded room fell silent as he began to speak: “I hope you all enjoyed your day off. While you were home doing nothing, I was busy thinking, and I’ve decided what must be done to get the LM program back on track.” He paused for dramatic effect, seeking eye contact with me and other program leaders.

  “This program is in terrible shape—right now it’s headed for failure. Clint Towl and Lew Evans have asked me to get personally involved to straighten things out, and that’s what I’m going to do. I’ve never failed at anything in my life, and I won’t allow LM to fail either.” He sniffed as his indignation heightened.

  “I’m very disappointed with the leadership of this program, and that goes all the way to the top. Can’t anyone on LM do anything right the first time, without alibis and excuses? You haven’t yet drawn up a schedule that held for more than a week. Every day we hear about new failures during tests, quality problems and just plain ‘goofs.’ What is the matter with all of you? What became of the Grumman spirit and talent you once had?

  Well, I’m going to set things right if I have to turn the LM program upside down and send all its current leaders packing.” Titterton’s body quivered with outrage as he delivered this threatening message.

  “And it’s not just me,” he went on. “NASA is very upset about Grumman’s poor performance on LM. Joe Shea told me he thought Grumman was complacent and technically arrogant—acting like ivory tower knuckleheads, I think he meant. Well, things are going to change, as of this moment. George Titterton is now in charge of LM, and George Titterton does not fail!”

  He then announced that I was being removed as LM Engineering manager and put in charge of the Manufacturing and Test operations in the Plant 5 LM final assembly hangar, with Howard Wright as my deputy. We would immediately locate to a command post in Plant 5 from which we would have authority to resolve all problems, whether in Engineering, Manufacturing, or Test, quickly and on the spot. It would be a two-shift, six-day-a-week operation with Wright and me working alternate shifts. Senior leaders from all the operating departments were also assigned to the command post and spacecraft team managers were appointed to lead a dedicated Assembly and Test team for each LM. All this was effective immediately, and the sooner we moved over to the command post the better.

  After the meeting Wright and I retreated in shock to my office. I had recently recruited Howard to join me on LM at the suggestion and assistance of Grant Hedrick. He was a creative electronic systems engineer who led the design and development of the complex avionics systems in the A6 Intruder and the E2 Hawkeye. Knowledgeable in digital systems and radar technology, he was gifted in assembling and integrating many individual systems elements into the architecture needed to perform major mission functions, such as weapons guidance or target detection and tracking. He devised algorithms and software to make system components operate in computer controlled harmony, and excelled in troubleshooting and debugging systems problems. He often flew on test flights aboard the airplanes to verify system performance and check newly developed fixes to problems. Sophisticated digital systems engineering that yielded practical results was his specialty, and his reputation within Grumman was soaring. I only succeeded in attracting him because Hedrick thought Wright had a capable avionics team backing him up on the airplane programs, and LM had a more urgent need for strong electronic systems engineering leadership.

  Howard Wright was tall and heavy-set, with large, round eyes, a high forehead, and thin, blondish hair. Careful and precise, he did not jump to conclusions until sure he had ferreted out all the pertinent facts and explanations. He had a wide network of associates throughout Grumman Engineering whose capabilities he had evaluated, and he freely drew upon them for help on LM regardless of which Grumman program they were working on. He was polite and optimistic and seemed to have no preconceptions about how space compared with aircraft engineering. Both required expertise, talent, attention to detail, and follow up in execution, which Wright provided in good measure. I was delighted to have him join me on LM—and embarrassed that he’d only been there a short while when Titterton had suddenly changed everything.

  We went to Joe Gavin’s office, where with his usual aplomb he calmed us down enough that we could start to plan how to carry out Titterton’s edicts. First order of business was to find a home for the command post. We began with a walk through the LM assembly hangar. It was a cavernous, environmentally clean room, about two hundred feet long by eighty feet wide, with thirty-five feet from the floor to the hook on the overhead traveling crane that ran the length of the ceiling. The room, the walls, ceiling, and floor of which were all white, was brightly lit by rows of recessed fluorescent lights in the ceiling. One long wall of the room was punctuated by large, fixed windows at first- and second-story heights through which activities in the assembly area could be viewed without entering the controlled environment. Both long walls were dominated by two-story LM assembly fixtures, movable steel structures that served as work scaffolds, providing ready access to all levels of the spacecraft. The assembly fixtures were also painted white, but they were heavily festooned with cables, plumbing, work carts, and support equipment of various sizes, which generally were black, gray, or brown. The LM under construction was barely visible when surrounded by its assembly fixture.

  To enter the assembly area we went into an adjacent locker room, where we donned medical-style white gowns, hats, and gloves, then brushed our shoes in a mechanical cleaner and donned nylon booties. Clear plastic safety glasses completed the attire required for all workers on the assembly floor. A huge blower and air filtration system, hidden behind the white walls, maintained a positive pressure in the entire assembly room so that dirt could not infiltrate inside. To enter we passed through a double set of pressure-sealed doors.

  There was no office space on the assembly floor, although there were some small desk areas that the quality inspectors used to process their paperwork. We quickly concluded that the nascent command post should not be on the assembly floor.

  Outside the clean room, the adjacent areas in Plant 5 housed the automated checkout equipment, which was provided and operated by General Electric for all Apollo prime contractors. There were three ACE control rooms at Grumman, which
allowed simultaneous electrical test and checkout of LMs in three of the assembly fixtures. The control rooms were large enough to house the numerous test engineers and quality inspectors needed to conduct rigorous, complex tests of the LMs. They had a row of test consoles on one wall, with cathode ray tube (CRT) screens providing readouts of the instrumentation data from the test vehicle. A wide range of stimuli and programmed sequences could be generated as required by the test plan, by the large mainframe computer which controlled ACE’s operations at the test conductor’s command. Two of the ACE rooms contained viewing windows overlooking the assembly floor, a feature which seemed desirable when the plans were laid out but was of limited practical benefit because of the difficulty of seeing anything within the cluttered assembly fixtures.

  With the adjacent Plant 5 real estate solidly occupied by ACE rooms and other indispensable equipment, our search for a home for the command post turned to a small group of office trailers in the parking lot of the courtyard behind the main building, near the outside entrance to the LM assembly entry locker room. These prefabricated buildings, known as the Nerve Center, had been set up a few months previously. They provided, in several spartan, low-ceilinged rooms, space in which the managers, foremen, and test teams could meet to discuss their plans and problems and forecast their schedules and requirements. It was, as a New York clothing chain once advertised, “no fancy fixtures—plain pipe racks.” Small windows, dark imitation wood paneling, and dirt-tracked vinyl tile floors created a drab, shabby environment. As we looked around without enthusiasm, Gavin, Wright, and I knew that this was where the command post must be. Although lacking in creature comforts, it was close to the action on the assembly floor but free of the cleanliness restrictions, and the trailer arrangement could be readily expanded and modified as our needs dictated. With its location settled, we began planning the staffing and activities of the command post.