Armstrong had begun to study the problem of how to land a flying machine on the Moon some seven and a half years before he became the commander of Apollo 11. “We knew that the lunar gravity was substantially different [roughly one-sixth that of Earth’s],” Armstrong recalls of the engineering work begun at Edwards following President Kennedy’s commitment in May 1961. “We knew that all our aerodynamic knowledge was not applicable in a vacuum. We knew that the flying characteristics of such a vehicle were going to be substantially different from anything we were accustomed to.”
The notion of attacking the unique stability and control problems of a machine flying in the absence of an atmosphere, through an entirely different gravity field, “That was a natural thing for us, because in-flight simulation was our thing at Edwards,” Armstrong relates. “We did lots and lots of in-flight simulations, trying to duplicate other vehicles, or duplicate trajectories, making something fly like something else.”
The assistant director of research at the Flight Research Center, Hubert M. “Jake” Drake, got the small group organized. Back in the early 1950s, Drake had played a similar catalytic role in conceptualizing ways to attain speeds of Mach 3 and altitudes over 100,000 feet in a research airplane, an initiative that led to the hypersonic X-15 program. Attacking the problem of a lunar landing research vehicle along with Drake were research engineers and frequent collaborators Gene Matranga, Donald Bellman, and Armstrong, the only test pilot involved.
The first idea that the Drake group considered was some form of helicopter, because of the helicopter’s abilities to hover and to take off and land vertically. Unfortunately, according to Neil, helicopters “could neither replicate the consequences of lunar gravity nor the handling characteristics of reaction system machines.”
Another idea that the Drake group entertained was to suspend a small lunar landing research vehicle beneath some sort of giant gantry and “fly” the vehicle tethered. Independently, a pair of researchers at NASA Langley in Virginia, Hewitt Phillips and Donald Hewes, later developed this idea into a useful simulator called the Lunar Landing Research Facility (LLRF). The mockup could not actually fly. Tethered to an overhead trestle, it moved in limited ways and was not the realistic flying machine that the FRC group sought. An even safer option was to go the route of an electronic, fixed-based simulator. Ultimately, NASA used all three methods—helicopters, Langley’s LLRF, and different electronic fixed-base simulators—to study the problems of lunar landing and to train Apollo astronauts.
Seeking to simulate as exactly as possible what Armstrong would later call flying “wingless on Luna,” Drake’s group opted for a bolder, more innovative scheme, one based on VTOL technology. VTOL referred to “vertical takeoff and landing.” It was a new and potentially revolutionary technology in which an aircraft equipped with translatable engines (like the Harrier jet that the British eventually built) flew with some helicopter-like traits.
“There were dozens of experimental VTOL machines during the late fifties and early sixties,” Armstrong relates, “and each of them had a unique attitude control system.” The person who was building the best-known VTOL test rigs at the time was British engineer A. A. Griffith. As Armstrong began his last semester at Purdue in 1954, stories appeared in the aviation and popular press about Griffith’s pioneering vertical takeoff and landing device. It was hard not to pay attention to the weird-looking machine. Its pilot sat in a control station atop an entirely open-air framework of tubing, a calliope of “puff pipes” for attitude control arranged all around him. The bizarre contraption earned the nickname the “Flying Bedstead.” Others called it the “Pipe Rack.”
“We were aware of that work, certainly to the extent that it was covered in Aviation Week,” Armstrong states. “However, since lunar gravity simulation was the foundation of our concept, and the British Flying Bedstead had no such system, the most value to us of such a craft was its reaction control system. An Earth-based VTOL had to be able to handle the winds, wind shears, and gusts of our atmosphere. A lunar landing machine had no such need. Of course, an Earth-based flying simulator would have some of those problems. That was our challenge—to build something simulating lunar conditions that could fly that way here on Earth.”
The Drake group’s only known contender for attitude control of a lunar flyer was a reaction system using small rockets. At the High-Speed Flight Station in the late 1950s, as mentioned in a previous chapter, researchers had devised a test rig known as the Iron Cross. The rig investigated the basic handling qualities and control needs of a reaction control system for use in the X-15. Unlike the Flying Bedstead, the Iron Cross did not actually fly, but it did employ nitrogen gas jets, which provided control moments for testing in simulated near-vacuum maneuvers.
The basic concept for a lunar landing research vehicle that the Drake group arrived at in mid-1961 was to mount a jet engine in a gimbal placed underneath the test vehicle so that the thrust produced by the jet always pointed upward. The jet would lift the test vehicle to the desired test altitude, whereupon the pilot would throttle back the engine to support five-sixths of the vehicle’s weight, simulating the Moon’s one-sixth gravity. The vehicle’s rate of descent and horizontal movement would be handled by firing two throttle-able hydrogen peroxide lift rockets. An array of smaller hydrogen peroxide thrusters would give the pilot attitude control in pitch, yaw, and roll. If the primary jet engine failed, auxiliary thrust rockets could take over the lift function, temporarily stabilizing the machine. What was so radical about the concept was that aerodynamics—the science on which all flying on Earth was done—played absolutely no part. In this sense, the lunar landing test vehicle that Armstrong helped to conceptualize in 1961 was the first flying machine ever designed for operation in the realm of another heavenly body, yet one that could also fly right here on Earth.
“Our first idea,” according to Armstrong, “was to have the mockup of the lander carried on another, larger vehicle and make that larger vehicle something that created the conditions that duplicated the lunar gravity and the lunar vacuum. Our thought was, when the actual vehicle got built—and at that point no one knew what the Apollo configuration would be—we could put something like it on top of this carrier and pilot-astronauts could fly it just like they would over the Moon. They could do it at Edwards or wherever, and learn how such a machine flew. Then we decided it was going to be a pretty complicated project, and that what we should do first was build a little one-man device that just investigated the qualities and requirements of flying in a lunar environment. With that, a database would grow from which we could build the bigger vehicle carrying the mockup of the real spacecraft.”
Through the summer and fall of 1961, the Drake team devised such a craft. According to Neil, “It looked like a big Campbell Soup can sitting on top of legs, with a gimbaled engine underneath it.”
Unknown to the Drake group, another team of engineers was also busy in late 1961 exploring the design of a free-flight lunar landing simulator. The news that this team worked at Bell Aerosystems in Buffalo, New York, came as no surprise. The descendant of the company that had built the X-1 and so many of the other early X-series aircraft, Bell Aerosystems was the only American aircraft manufacturer with any significant experience in the design and construction of VTOL aircraft using jet lift for takeoff and landing. Jake Drake heard about the Bell initiative from a NASA Headquarters official in the fall of 1961. “We’ve just had a proposal from some people at Bell for a machine to do what you’re talking about doing,” the official told Drake. “You ought to go talk to them.” According to Gene Matranga, “We talked to them, and they had not only thought about, they were much further down the road to a practical solution to the problem.”
Immediately Drake invited Kenneth L. Levin and John Ryken, two of the principal Bell engineers at work on the concept (a third was John G. Allen Jr.), to Edwards for consultation. Subsequently, Bellman and Matranga traveled to Bell, where they rode the company’s Model 47 helicopters on simulated lunar descents. Armstrong did not make the trip, because he had heavy responsibilities at the time in the X-15 and Dyna-Soar programs. It was also shortly before his daughter died. What the FRC engineers saw at Bell confirmed their strong suspicion that helicopters just could not fly the descent trajectories and sink rates that came close to what was expected for a lunar lander. Helicopters could approximate a variety of final descent trajectories, but to do that often required their flying for substantial periods inside the so-called Dead Man’s Curve, the terminal phase of a descent trajectory where it would be impossible to abort safely without crashing into the surface.
NASA contracted with Bell to draw up blueprints for a small, relatively inexpensive lunar landing test vehicle whose design would be independent of the actual Apollo configuration, which it had to be, since the Apollo configuration had not yet been decided upon. Bell’s job was to lay out a machine with which NASA could investigate the inherent problems of lunar descent from altitudes up to 2,000 feet with vertical velocities of up to 200 feet per second. Results from the $2.5-million study, NASA felt, could significantly help in the design of the Apollo spacecraft, a much larger contract that North American Aviation, Inc., had been awarded the previous autumn.
Not until July 1962 did NASA settle on how to go to the Moon. When JFK boldly called for the Moon landing, a great many qualified engineers and scientists envisioned getting there and back in one brute rocket ship. That was how Jules Verne and most other visionaries had seen it happening. A gargantuan rocket roughly the size—and no doubt the weight—of the Empire State Building would take off from Earth, fly to the Moon, back down rear end first to a landing, and blast off for home. It was a mission mode that advocates called Direct Ascent. A new rocket with twelve million pounds of thrust, by far the most powerful booster ever built, would take astronauts directly from the Earth to the Moon, with no stops in between. The name of the proposed monster was the Nova.
A second major option for the lunar landing—and one that many spaceflight experts, including NASA’s rocketmeister Dr. Wernher von Braun, came to favor—was Earth Orbit Rendezvous, or EOR. According to this plan, a number of the smaller Saturn-class boosters being designed by the von Braun team at Marshall Space Flight Center in Alabama would launch components of the to-be-lunar-bound spacecraft into Earth orbit, where those parts would be assembled and fueled for a trip to the Moon and back. The main advantage of EOR was that it required far less complicated booster rockets, ones nearly ready to fly. Just two or three of von Braun’s early Saturns would do the job. Another benefit of EOR could have been long-term: in the process of going to the Moon, the U.S. space program might have built a platform in Earth orbit that could be converted into a space station.
To the surprise of many experts, NASA selected neither Direct Ascent nor Earth Orbit Rendezvous. On July 11, 1962, officials announced that a concept known as Lunar Orbit Rendezvous would be America’s way to the Moon. Lunar Orbit Rendezvous, or LOR, was the only mission mode under consideration that called for a customized lunar excursion module to make the landing.
The LOR decision was made over the strenuous objections of President Kennedy’s science adviser, Dr. Jerome Wiesner. Like other skeptics, Wiesner felt that LOR was too risky to try. If rendezvous had to be part of the lunar mission, he felt that it should be attempted only in Earth orbit. If rendezvous failed there, the threatened astronauts could be brought home simply by allowing the orbit of their spacecraft to decay. If a rendezvous around the Moon failed, the astronauts would be too far away to be saved. Nothing could be done. The specter of dead astronauts sailing around the Moon haunted those who were responsible for the Apollo program and made objective evaluation of its merits unusually difficult.
In the end, NASA’s mission planners determined that LOR was no more dangerous than the other two schemes, likely even less dangerous, and that it enjoyed several critical advantages. It required less fuel, only half the payload, and somewhat less new technology. It did not require the monstrous Nova, and it called for only one launch from Earth, whereas the once-favored EOR required at least two. Trying to bring down a behemoth like the upper stage of a Nova onto the cratered lunar surface would be next to impossible, as every analysis came to show. Even if a landing with Nova could somehow be managed, there would still be the problem of the astronauts getting down to the lunar surface from atop such a giant structure, for, even after all of its rocket staging, the spacecraft that landed would still be about the size of the Washington Monument. Engineers had even looked into the design of a transport elevator for the spacecraft for that purpose. A Moon landing via EOR looked only marginally easier. The ship leaving for the Moon after Earth orbit rendezvous would be smaller than the battleship-sized Nova, but it would still be a very ponderous stack of machinery to eyeball down to a pinpoint landing. After months of study, with absolutely no satisfactory answers to the landing dilemmas of Direct Ascent or EOR appearing, there was no choice but to go with LOR.
The greatest technological advantage of LOR was that it turned the lander into a “module.” Only the small, lightweight lunar module (LM), not the entire Apollo spacecraft, would have to land on the Moon. Also, because the lander was to be discarded after use and would not be needed to return to Earth, NASA could customize the LM’s design for maneuvering flight in the lunar environment and for a soft, controlled lunar landing, and for nothing else. In fact, the beauty of LOR was that NASA could tailor all of the modules of the Apollo spacecraft independently—the command module (CM), service module (SM), and LM. The modularity extended to the LM itself. It would be a two-stage vehicle. The entire LM would descend to the surface using a throttle-able rocket engine. But the lower module, holding the landing legs, descent engine, and associated fuel tanks, would remain on the lunar surface and act as the launch platform for the upper or ascent stage, with its separate fixed-thrust engine, associated tankage, attitude control rockets, and, of course, cockpit.
Most important, LOR was the only mission mode by which the Moon landing could be achieved by Kennedy’s deadline of decade’s end. For NASA, that was the clincher. The phrase Armstrong remembers is that “LOR saves two years and two billion dollars.”
The promise of the preliminary LLRV design played a very small but not inconsequential role in the LOR decision. The key people making the decision in favor of LOR at NASA were Associate Administrator Robert C. Seamans Jr.; Brainerd Holmes, the head of the Office of Manned Space Flight; George Low, Holmes’s director of spacecraft and flight missions; and Joseph F. Shea, head of the Office of Manned Space Flight Systems. They made the decision in personal consultation with Bob Gilruth, director of the Manned Spacecraft Center, and Wernher von Braun, director of the Marshall Space Flight Center.
Von Braun’s preference for LOR had surprised his own staff. At the end of a daylong briefing given to Joe Shea at NASA Marshall on June 7, 1962, the immigrant German rocketeer had announced, “We at the Marshall Space Flight Center readily admit that when first exposed to the proposal of the Lunar Orbit Rendezvous Mode we were a bit skeptical—particularly of the aspect of having the astronauts execute a complicated rendezvous maneuver at a distance of 240,000 miles from the Earth where any rescue possibility appeared remote. In the meantime, however, we have spent a great deal of time and effort studying the [different modes], and we have come to the conclusion that this particular disadvantage is far outweighed by [its] advantages.”
Overnight, a landing module became one of the most critical systems, if not the most critical system, in the entire Apollo program. The big Saturn V rocket could propel astronauts inside their snug command module into lunar orbit, but unless a special lander went along also, there would be no way for them to land. And Apollo was all about landing.
Immediately, serious work on the LM began. In November 1962, the Grumman Corporation of Long Island, New York, won the contract. The evolutionary path to a finished LM turned out to be torturous. The cabin volume was changed from spherical to cylindrical. The landing legs were reduced in number from five to four. The window area was substantially reduced. The seats were removed, meaning that the two-man crew of the LM would stand, like trolley conductors. Not only did the new standing arrangement move the pilots’ eyes closer to the “windshield,” improving visibility, it also reduced the weight of the overall structure, which was a crucial factor in everything related to the LM design. But it also meant that a cable restraint system had to be devised to keep the pilots in the proper position inside the cabin and hold them secure during the impact of touchdown.
A long string of test failures—propulsion leaks, ascent-engine instabilities, stress corrosion of various aluminum alloy parts, electric battery problems—kept the Grumman team busy fixing and refining its extraordinary machine for nearly seven years. Not until March 1969 was even the first LM ready to test-fly. It took place in Earth orbit, as the primary task of Apollo 9. Throughout most of its developmental life, everyone called the vehicle the “LEM,” until May 1966 when a memo from the NASA Project Designation Committee officiously changed the name simply to “LM.” Apparently, the word “excursion” sounded too much like a vacation rather than a deeply serious enterprise for human space exploration. In the vernacular, people still pronounced the acronym, not as two individual letters, but as if the vowel were still there.
With the LOR decision in hand, the requirements for the Flight Research Center’s lunar landing research vehicle became much more explicit. Strictly by chance, the characteristics, size, and inertias of the original LLRV design were very much like what Grumman soon realized it needed to build into the LM. “Bell already had a design for the LLRV,” Gene Matranga relates, “but went through a very quick redesign when the concept of the lunar landing changed to LOR. As it turned out, the revised machine wound up being a better solution to the problem.”
Bell Aerosystems began fabricating two LLRVs (of the same exact design) in February 1963; NASA wanted the vehicles ready in sixteen months’ time. On April 15, 1964, the machines arrived at Edwards, as requested, disassembled and in boxes, because FRC technicians wanted to install their own research instruments and believed they could complete the craft more expeditiously than could Bell. President Lyndon Johnson saw an assembled LLRV on a visit to Edwards in mid-June 1963; the politician must have chuckled at its Rube Goldberg appearance. Standing ten feet tall and weighing 3,700 pounds, the LLRV had four aluminum truss legs that spread out across some thirteen feet. The pilot sat out in the open air, behind a Plexiglas shield. He sat in a specially designed rocket ejection seat built by Weber Aircraft, one of the least known of the American ejection seat manufacturers, yet one of the largest. Weber’s seat was so effective that it operated successfully even at “zero-zero,” the lowest point in an ejection envelope, and could do so safely even if the LLRV was moving downward at a rate as high as thirty feet per second. No ejection seat ever performed better, which was a good thing given that it would have to be used more than once in the LLRV program.
The first pilot to fly the LLRV was Neil’s former boss Joe Walker. Walker made the inaugural flight on October 30, 1964, the day after astronaut Ted Freeman’s fatal accident outside Houston in his T-38 trainer. The inaugural flight consisted of three brief takeoffs and landings totaling just under a minute of flight time. Prudently, Walker took the machine no higher than ten feet and used only the main jet (a General Electric CF-700-2V turbofan engine producing 4,200 pounds of thrust) for lift. He did not activate the two lift rockets, but he did briefly fire all sixteen of the small hydrogen peroxide control rockets (grouped in pairs of two) for attitude control. Walker compared taking off in the machine to rising up in an elevator, except for the hissing sound produced by the strange and grotesque piece of hardware when he fired short bursts of the reaction controls. When that happened, a cloud of peroxide steam nearly enveloped the craft, giving rise to another nickname, “the belching spider.” Fellow FRC test pilot Donald Mallick and Emil “Jack” Kleuver, an army test pilot on loan to NASA, later flew the machine, Mallick making the most LLRV flights of all, over seventy.
Between 1964 and the end of the LLRV test program in late 1966, some 200 research flights were carried out at Edwards. Pilots could operate the vehicle in one of two modes. They could fly it as a “conventional” VTOL with the jet engine locked in position and providing all the lift; pilots called this the “Earth mode.” Or they could fly it in the “lunar mode” in which the engine could be adjusted in flight to reduce the apparent weight of the LLRV to its lunar equivalent. In the lunar mode, as stated earlier, lift was provided by a pair of controllable 500-pound-thrust rockets (noncombustion rockets using a 90 percent solution of hydrogen peroxide as fuel) that were fixed to the fuselage outside the gimbal ring. Operating in the lunar mode, the pilot could modulate the angle and thrust of the engine to compensate for aerodynamic drag in all axes. Generally, the pilots preferred flying the Earth, or VTOL, mode. As Armstrong notes, “In the lunar simulation mode, uncomfortably large attitudes were required for reasonable decelerations.” On the other hand, the sensitive throttle for the rocket engine made altitude control much better in lunar simulation.
As strange as it all was, the LLRV compensated for its Earthbound existence in some very clever ways and closely duplicated what it was like to fly over the Moon, though its highest altitude reached just under 800 feet, and its longest flight lasted something less than nine and half minutes. Amazingly, no serious accidents occurred during the entirety of the Flight Research Center’s LLRV program.
Armstrong had left Edwards for Houston in September 1962, so he was unable to stay as informed about the LLRV program as he would have liked. “I did go to Edwards a few times and talked with Joe Walker. I was aware of some of the difficulties they were encountering in developing a satisfactory flight control system for the vehicle. I would have liked to have been more involved, but I was loaded with other responsibilities at the time.” Gene Matranga confirms that Neil stayed a part of the LLRV program. “Neil got tabbed by the Houston people to be the engineering pilot focal point,” he remembers, “making sure that the things we were doing met the needs of the astronauts.” But Neil made it to Edwards only once to see the machine fly. NASA did not want him or any of the other astronauts to fly the contrary and risky machine. Still, as Matranga relates, “Neil made sure that we were doing the things that Houston wanted us to do.”
Ground simulators offered considerable help. As Neil explains, “Traversing large pitch or roll angles required more time or larger control power. It was expected that control characteristics ideal on Earth might be not at all acceptable on the Moon.” As a result of hundreds of hours in the simulators, the astronauts found that good control could be obtained with “on-off” rockets that had been mechanized for rate command—that is, for the vehicle’s angular rate (or rate of change) proportional to control deflection—but they were still, in Neil’s words, having “some difficulty in making precise landings and eliminating residual velocities at touchdown, probably due to a pilot’s natural reluctance to make large attitude changes at low altitudes.”
The Lunar Landing Research Facility at Langley was an imposing 250-foot-high, 400-foot-long gantry structure that had become operational in June 1965 at a cost of nearly $4 million. Armstrong considered the LLRF “an engineer’s delight.” “It worked surprisingly well,” says Armstrong. “The flying volume—180 feet high, 360 feet long, and 42 feet wide—was limiting, but adequate to give pilots a substantive introduction to lunar flight characteristics.” To make the simulated landings more authentic, its Langley designers filled the base of the huge eight-legged, red-and-white structure with dirt and modeled it to resemble the Moon’s surface. Often testing at night, they erected floodlights at the proper angles to simulate lunar light and installed a black screen at the far end of the gantry to mimic the airless lunar “sky.” Technicians climbed into the fake craters and sprayed them with black enamel so that the astronauts could experience the shadows that they would see during the actual Moon landing.
Though “the engineers at Langley did some wonderful work trying to create a flexible [cable and pulley] system that allowed it to feel like a real flying spacecraft,” control for pitch and roll could be, in Neil’s words, “excessively sluggish.” “The LLRF was a clever device,” in Armstrong’s judgment. “You could do things in it that you would not want to try in a free-flying vehicle, because you could be saved from yourself.”
In 1964, the Astronaut Office looked around to see what VTOL machines might be available as possible lunar landing simulators. Deke Slayton asked Armstrong specifically to look into the potential of the Bell X-14A. This was a small and versatile aircraft that employed the same vectored-thrust and reaction-control arrangement used by the British Harrier. Houston knew that engineers at NASA’s Ames Research Center in Northern California were using the X-14A to simulate lunar descent trajectories, so Armstrong flew out for a visit. In February 1964, he made ten evaluation flights to see if the X-14A had any applicability to lunar landing simulations. Neil concluded that, while a pilot could simulate a lunar trajectory in the X-14A, the attitude changes required were that of an Earth-gravity VTOL machine and could not replicate lunar motions. In that sense, it flew more like a helicopter than it did a lunar module. The X-14A also had a problem with ground effects. When a helicopter descended toward the ground in hovering flight, the amount of power it required to stay aloft got smaller; however, with the X-14A (and many other VTOLs) the effect was the reverse. The closer to the ground it got, the more throttle it took. Reingestion of hot exhaust gases near the surface caused disconcerting instabilities and a reduction in thrust. Also, so much movement in the throttle developed in the final phase of descent that smooth touchdowns were a rarity. “It was hoped and expected that the actual lunar lander would have little ground effect,” Armstrong explains, “somewhere between the helicopter and the VTOL.” For that degree of accurate simulation, another class of training vehicle was required.
“Having no flying machines to simulate lunar control characteristics was frustrating the Astronaut Office,” Armstrong recalls. The only effective alternative was to try the Flight Research Center’s LLRV, however risky some people in NASA considered the highly unusual free-flight vehicle. Heading the program at Houston to convert the LLRV into an astronaut trainer was Dick Day, the simulations expert from the Flight Research Center who back in 1962 had helped Neil to become an astronaut.
The decision to turn the LLRV into a trainer, or LLTV, came early in 1966, just prior to Armstrong’s Gemini VIII flight. By this time, Grumman had come a long way to finalizing the design of the LM, whose first test flight, designated Apollo 5, was scheduled for January 1968. (Apollo 9 tested the lunar module in March 1969, the first manned test of the LM.) Building an actual LM that could fly on Earth like the LLRV was possible but, according to Armstrong, would have been “prohibitively time consuming and expensive.” As it was, the LLRV, although it predated the LM by five years, was not all that different in physical size and control rocket geometry from what had become Grumman’s actual vehicle. Relatively quickly and inexpensively, NASA got Bell to produce an advanced version of the LLRV that even more closely matched the characteristics of the LM.
The decision to build LLTVs brought Neil back squarely into lunar landing studies. In the summer of 1966, as he was preparing for his backup role in Gemini XI, Houston ordered three LLTVs at a cost of roughly $2.5 million each. At the same time, the Manned Spacecraft Center requested that the Flight Research Center prepare its two LLRVs for shipping to Houston as soon as the FRC engineers were done with them. Even before he left on the Latin America tour, Neil participated in discussions with Bell on what was needed in the LLTV design. With MSC test pilot Joseph S. Algranti, he went to Edwards in August 1966 to check out the LLRV. (Algranti had checked out in the LLRV several months earlier.) Although he did not fly the machine during that visit, Neil did fly LM trajectories in a helicopter with Algranti. Upon returning from the Latin American tour, he became routinely involved in LLTV matters. He was on the scene when LLRV number one arrived in Houston from Edwards on December 12, 1966. When FRC test pilot Jack Kleuver came to Houston to verify that the machine was working, Armstrong observed. When Algranti and his fellow MSC test pilot H. E. “Bud” Ream made the first familiarization flights with it at Ellington AFB near the Manned Spacecraft Center, Neil watched the operation and studied their ground rules. He spent January 5 to 7, 1967, with Algranti in Buffalo, participating in the LLTV Design Engineering Inspection at Bell. A few days later, he and Algranti were off to Edwards to review the final results of the LLRV program. While in California, Neil flew some LM trajectories in a Bell H-13 helicopter. He also witnessed an LLRV flight piloted by Jack Kleuver. Immediately after attending the funerals for the Apollo 1 crew at Arlington National Cemetery and West Point in late January, Armstrong and Buzz Aldrin flew in a T-38 directly to Langley Field in order to make simulated lunar landings on the LLRF. It was Neil’s first time on Langley’s gadget and it would not be his last. On February 7, 1967, he and Buzz flew a T-38 to Los Angeles to be custom-fitted for an LLRV ejection seat at Weber Aircraft. Later in the month, he went again to Los Angeles, this time to North American (and with Bill Anders), to review the design for the tunnel through which the astronauts would move back and forth between the Apollo command and service module (CSM) and the LM. In March 1967, he traveled to the West Coast once more, to Los Angeles and to San Diego, where he reviewed the LM landing radar program at Ryan Aircraft. During these months, he also got in a good bit of helicopter time in order to get ready for training in the LLTV. Little wonder that when NASA assembled its Apollo fire investigation panel, Armstrong was nowhere in the picture. He was too deeply immersed in matters related to lunar landings.
Helping transform the research vehicle into a training vehicle was a challenge for which Armstrong as an engineer, test pilot, and astronaut was extremely well suited. Back in 1961, he had contributed to the machine’s original concept. Bell built the LLTV essentially on the same structure as the LLRV, but now the main goal was to replicate as closely as possible the trajectory and control systems of the LM. Certain flying characteristics of the LM could not be replicated, however. Most notably, it was impractical, if not impossible, to design the LLTV so that it provided the rate of descent that the LM had.
Another goal was to make the LLTV as much like the LM in terms of critical design features. For example, Bell built the new LLTVs with an enclosed cockpit that enjoyed LM-like visibility. To match the LM configuration, it also moved the control panel from the center of the cockpit to the right side and set up the same array of visual displays. The LLTV was given a three-axis sidearm control stick very comparable to what Grumman was placing into the LM (rather than the conventional aircraft-type center stick for pitch and roll control and rudder pedals for yaw control that had been in the LLRV), and a rate-command/attitude-hold control system closely approximating the handling characteristics anticipated for the LM. The LLTV also incorporated a compensation system that sensed any aerodynamically induced forces and moments and provided automatic correction through the engine and attitude rocket system. In this way, the motions of the LLTV even more closely approximated flight in a vacuum. Several improvements were made in the electronics system to take advantage of the same miniaturized, lightweight components that were being used in the LM. Other improvements included an improved ejection seat, more peroxide for the rockets to increase their duration, a slightly upgraded jet engine, and a modified attitude to be more like the LM.
Armstrong was involved in the LLTV Design Engineering Inspection at Bell, so, in his words, he “must have more or less agreed with all the Bell proposals for the Bell LLTV—at least I had the chance to give my input.” In Neil’s view, “it was very necessary to have the LLTV control system replicate the LM. It was not necessary, nor was it attempted, to provide simulations of the LM environmental systems, communications systems, guidance systems, et cetera. The LLRV had had a number of system reliability and component problems, and many of the LLTV changes were intended to improve those areas.”
Not all the changes from the LLRV to the LLTV were universally considered to be improvements, at least not by the team of FRC engineers who had made the LLRV program so successful at Edwards. In trying to make the training vehicle as much like the LM as possible, Gene Matranga asserts that Bell eventually added some systems that actually made the machine less reliable to fly. Most notably, Bell changed from an analog “fly-by-wire” (FBW) control system to a digital system, because that was the type being used in the LM. Unfortunately, “dead periods” existed within the circuitry of the digital system during which the pilot could not sense the loss of electrical power. In January 1971, just such an electrical system failure caused NASA test pilot Stuart Present to lose control (the switchover to battery backup power did not work), forcing him to eject and the LLTV to crash hard into the ground at Ellington, destroying the vehicle.
Not just the three new LLTVs were used for astronaut training; so too were the two older LLRVs. The Manned Spacecraft Center modified the existing machines into trainers, dubbing them LLTV A1 and LLTV A2. The three new machines, the first of which arrived from the Bell factory in December 1967, became LLTV B1, B2, and B3. Before the astronauts were allowed to fly any of them, they received a couple months’ flight instruction from Joe Algranti and Bud Ream, the MSC test pilots who had gone out to Edwards to learn how to “master” the LLRV. The astronauts that Slayton designated as potential LM crewmen, including Armstrong, then went to helicopter school for three weeks, to Langley’s LLRF for a week, and finally to fifteen hours in a ground simulator before they got their first chance to fly an LLTV, always at nearby Ellington. As Neil had already gone to navy helicopter school in 1963 and had built up quite a bit of “helo” time over the next four years, all he had to do as far as helicopters were concerned was brush up on his skills prior to his LLRV checkout.
“The helicopter wasn’t a good simulation of the lunar module control at all,” Armstrong explains. “Had it been, we would have configured a helicopter such that it could duplicate lunar flying. That could have been done with a great deal less risk than flying the LLRV or LLTV. But we never could come up with anything that worked at all well. The natural requirements of helicopter aerodynamics preclude you from duplicating the lunar module characteristics. Nevertheless, the helicopter was valuable to understand the trajectories and visual fields and the rates. You could precisely duplicate the flight paths that you wanted. It’s just that the control you were using to do that was not at all the same.”
Astronaut Bill Anders, who also flew the LLTV several times, wonders in retrospect why NASA did not think through its helicopter training for the astronauts more thoroughly: “That was, in a sense, almost bad training, flying helicopters. If you had a helicopter on the Moon, you would be fooled at first because the helicopter’s mass would be the same but it would have one-sixth the weight and one-sixth the lift. When you tilted it up, it would have one-sixth the retarding force, so therefore you are probably going to go six times beyond the landing point. Flying on the Moon was literally a different world.”
Frank Borman and Jim Lovell agreed. Borman, who flew the LLTV only once, on May 6, 1968, called flying it “a hairy deal.” “From my standpoint it was a dicey training aid.” Lovell: “Even though the LLTV and a helicopter operated entirely opposite in terms of controls, they still had you doing all this practicing in helicopters. Add to that all of the different safety factors you had to worry about—the ejection seat, the throttle, how to fly the thing, go up to two hundred feet and have two minutes of fuel and that was it. It made me worry about flying it.”
An important thing to say about Armstrong, in the view of both Anders and Lovell, is that Neil was the kind of experienced engineering test pilot who did an outstanding job thinking through what it took to fly in the unusual lunar environment and not letting his helicopter training dominate his piloting decisions. “Neil’s first experiences in the LLTV were probably not typical of the rest of us,” Anders relates. “It’s not that he had less helicopter time. It’s that Neil always thought these things through. If it required something counterintuitive or otherwise against the grain, he figured it out. The guys who flew more intuitively or who relied too much on their helicopter experience would have tumbled into craters or landed on rocks if they had tried to land on the Moon. In my view, the LLTV was a much undersung hero of the Apollo program.”
Eventually, all prime and backup commanders of Apollo lunar landing missions practiced on the LLTV. As the program went on, there was not enough LLTV time available and the backup commanders were cut short. The commanders usually flew a total of twenty-two flights, the backup commanders maybe a dozen. The astronauts who flew the LLTV besides Armstrong were Borman, Anders, Conrad, Scott, Lovell, Young, Shepard, Cernan, Gordon, and Haise.
Neil’s initial LLTV flight came on March 27, 1967, when the machine first came to Ellington Field; he made two flights in LLTV A1 that day. Algranti and Ream also flew the LLTV that month, but due to a combination of technical problems the machine was not flown for the remainder of the year. (None of the three new LLTVs were ready for flight testing until the summer of 1968.) When the LLTV came back on line, Armstrong was the first to get checked out in the machine (again, LLTV A1), followed by checkouts for Anders, Conrad, and Borman. Neil’s logbooks show that, between March 27, 1968, and April 25, 1968, he made ten flights in the converted LLRV. In Neil’s opinion, “The LLTV proved to be an excellent simulator and was highly regarded by the astronauts as necessary to lunar landing preparation.”
Yet the LLTV was also a highly dangerous machine to fly. “Without wings,” as Buzz Aldrin has noted, “it could not glide to a safe landing if the main engine or the thrusters failed. And to train on it properly, an astronaut had to fly at altitudes up to 500 feet. At that height a glitch could be fatal.” Armstrong found out just how unforgiving the vexatious machine could be on Monday afternoon, May 6, 1968, just fourteen months before the Apollo 11 landing.
“I wouldn’t call it routine, because nothing with an LLTV was routine, but I was making typical landing trajectories during the flight that afternoon, and as I approached the final phase of one of them, in the final 100 feet of descent going into landing, I noted that my control was degrading. Quickly, control was nonexistent. The vehicle began to turn. We had no secondary control system that we could energize—no emergency system with which we could recover control. So it became obvious as the aircraft reached thirty degrees of banking that I wasn’t going to be able to stop it. I had a very limited time left to escape the vehicle, so I ejected, using the rocket-powered seat. The ejection was somewhere over fifty feet of altitude, pretty low, but the rocket propelled me up fairly high. The vehicle crashed first, and I drifted in the parachute away from the flames and dropped successfully in the middle of a patch of weeds out in the center of Ellington Air Force Base.”
During the explosive ejection, the first he had experienced since abandoning his crippled Panther jet over Korea seventeen years earlier, Neil accidentally bit hard into his tongue. That was his only injury, except for a bad case of chiggers from the weeds. “It’s hard to compare against combat when a big shell from an aircraft just misses you. That’s close too, but the LLRV accident was indeed one of the close ones.”
To NASA Headquarters, Houston sent the following priority telegram about the accident:
LLRV #1 CRASHED MAY 6, 1968, AT 1328 CDT AT EAFB, TEXAS. PILOT, NEIL A. ARMSTRONG, NASA.MSC ASTRONAUT, EJECTED AFTER APPARENT LOSS OF CONTROL. ARMSTRONG INCURRED MINOR LACERATION TO TONGUE, VEHICLE WAS ON STANDARD LUNAR LANDING TRAINING MISSION. ESTIMATED ALTITUDE AT TIME OF EJECTION 200 FEET. LLRV #1 TOTAL LOSS—FIRST ESTIMATE $1.5 MILLION. PROBABLE CAUSE—NOT KNOWN AT THIS TIME. PROGRAM DELAY PROBABLE. LLRV #2 WILL NOT COMMENCE FLIGHT STATUS UNTIL ACCIDENT INVESTIGATION HAS BEEN COMPLETED AND CAUSE DETERMINED (LLTV’S HAVE NOT COMPLETED GROUND TEST PHASE AND THEREFORE ARE NOT APPLICABLE). BOARD OF INVESTIGATION APPOINTED BY DIRECTOR MSC.
Chairing the accident investigation board was Joe Algranti. Serving with him were Bill Anders and Pete Conrad, both of whom had also begun to fly the LLTV. (Conrad was a member only temporarily, until replaced by the FRC’s Don Mallick.) The statement in the telegram that Neil’s ejection took place from an estimated height of 200 feet was either in error or purposefully exaggerated to soothe fears in Washington about the dangers of flying the vehicle.
Those who observed the accident or who subsequently heard about it felt that Armstrong was very lucky to be alive. According to Chris Kraft, the frightening films he watched of the accident showed that Neil escaped death by just two-fifths of a second. “Winds were gusting that day,” Kraft describes, “something that can’t happen on the airless Moon, but Armstrong was fully in control for the first five minutes. He took it up several hundred feet and was ready to practice a nearly vertical descent and landing. Then the machine suddenly dropped. He steadied it and climbed back up another two hundred feet. Then the LLTV began to bounce around in the sky. It pitched down, then up, then sideways. Its stabilization had failed and it was clearly out of control. A ground controller radioed Neil to bail out. He activated the ejection seat with only a fractional second of margin. Neil’s parachute opened just before he hit the ground. He wasn’t hurt, but the LLTV was demolished in a fireball.”
Buzz Aldrin was not at Ellington, either, to see the accident, but he, too, understood it to be a near fatality: “When the machine began to wobble and spin during [Neil’s] descent from 210 feet to the runway, he fought to regain control with the thrusters, but the platform sagged badly to one side and lurched into a spin. He had maybe a second to decide. If the trainer had tipped completely over and he had fired his ejection seat, the rocket charge would have propelled him headfirst into the concrete below. Neil held on as long as he could, not wanting to abandon an expensive piece of hardware. At the last possible moment, he realized the thruster system had completely malfunctioned, and he pulled his ejection handles. He was blasted up several hundred feet, and his parachute opened just before he struck the grass at the side of the runway. Neil was shaken up pretty badly, and the LLRV exploded on impact.”
The cause of the accident was a poorly designed thruster system that allowed Armstrong’s propellant to leak out. Loss of helium pressure in the propellant tanks caused the attitude rockets to shut down, producing loss of control. “There was very little time to analyze alternatives at that point,” Neil explains. “It was just because I was so close to the ground. So, again, it was a time when you had to make a quick decision. You ‘departed.’ ”
The fact that NASA was flying the vehicle in such windy conditions was a major contributing factor. At Edwards, the FRC engineers had put a fifteen-knot limit on wind speed for LLRV flying, but the Houston staff felt they had to raise it to thirty knots in order to be able to use the machine on a regular basis. The FRC’s Gene Matranga feels that this is what really got Neil in trouble: “That afternoon there was a higher wind than we normally dealt with. Neil was using more attitude control fuel than had been budgeted because of the higher winds and the attitude rockets were firing much more continuously. The way we had designed the system, the fuel would go down so far, and then there was a standpipe that allowed you to have fuel saved for the lift rockets in case you had to use the lift rockets to recover the vehicle if the jet engine failed. The gentleman at Ellington who was responsible for watching the fuel consumption apparently froze at the switch. Neil should have shut off the lift rockets and saved his fuel for the attitude rockets and go back to the jet engine. Nobody warned Neil of that. What happened was the helium pressure, which sat on the top of the fuel to pressurize the system, was expended very rapidly, and Neil wound up having fuel but no pressurizing gas, because the lift rockets were in the open position. In essence, he had no control. He had to jump out.”
Interestingly, back before the LLRVs came to Houston, Armstrong had been present at Edwards to observe ground tests in which the FRC engineers, in association with technicians from Weber Aircraft, tested the lightweight ejection seat (less than 100 pounds) at different angles. Neil watched as the seat fired a human-sized dummy into the air and then saw that dummy smash hard into the ground after too few swings on its parachute. “Neil didn’t think much of that,” Matranga laughs. “He wasn’t terribly thrilled. But, as it turned out, that seat saved his life.”
After his accident on May 6, 1968, Armstrong behaved, typically, as if absolutely nothing out of the ordinary had just happened. Upon returning from a late lunch, astronaut Al Bean returned to the Astronaut Office and saw Neil at work at his desk in the office the two men shared. A little later, Bean went out in the hallway and walked over to a group of colleagues who were talking; he thought he heard them say that somebody had just crashed the LLTV. According to Bean, “I’m saying, ‘What happened?’ and they said, ‘Well, the wind was high and Neil ran out of fuel and bailed out at the last minute and the ejection seat worked and he lived through it.’ I said, ‘When did this happen?’ They said, ‘It just happened an hour ago.’ ‘An hour ago!’ I said, ‘That’s bullshit! I just came out of my office and Neil’s there at his desk. He’s in his flight suit, but he’s in there shuffling some papers.’ And they said, ‘No, it was Neil.’ I said, ‘Wait a minute!’ So I go back in the office. Neil looked up and I said, ‘I just heard the funniest story!’ He said, ‘What?’ I said, ‘I heard that you bailed out of the LLTV an hour ago.’ He thought a second and said, ‘Yeah, I did.’ I said, ‘What happened?’ He said, ‘I lost control and had to bail out of the darn thing.’ ”
Bean continues his story: “So, I went back down and told the other guys. Nobody gave a thought to rush down there and ask Neil about it. If it had been Pete Conrad, everybody would have rushed down there, because Pete would have regaled them with a great story. I don’t think it was that Neil was so extraordinarily cooler than the other guys. But, offhand, I can’t think of another person, let alone another astronaut, who would have just gone back to his office after ejecting a fraction of a second before getting killed. He never got up at an all-pilots meeting and told us anything about it. That was an incident that colored my opinion about Neil ever since. He was so different than other people.”
Neil’s reaction to hearing Al Bean’s story is just as circumspect as his behavior in the story itself: “That is true, I did go back to the office. I mean, what are you going to do? It’s one of those sad days when you lose a machine.”
Once more, as had been the case in his Gemini VIII flight, Armstrong rightfully came out of the experience with an enhanced reputation for being able to handle an emergency situation—and this time there was no Monday morning quarterbacking.
“The LLTV was widely regarded—and properly so—as a high-risk vehicle,” Armstrong admits, “and one with which the management felt very uncomfortable. But the pilots universally, although they may have not liked the vehicle or liked flying it, they all agreed that it was the best simulation we had and gave us by far the highest confidence about what it was like to fly in the lunar environment.”
Houston grounded the LLTV pending the findings not only of the MSC’s accident investigation team but also of a special review board appointed by Dr. Thomas O. Paine, the man who succeeded James Webb as NASA administrator in late 1968 following the election of Richard M. Nixon. Chaired by General Samuel C. Phillips, the chief manager of the Apollo program under Dr. George Mueller, the committee at headquarters studied the impact of Armstrong’s crash on the overall Apollo program, particularly the LM. By mid-October 1968, the two reports were out, urging LLTV design and management improvements, yet clearing the program to continue.
Four minutes into a planned six-minute flight on December 8, 1968, MSC chief test pilot Joe Algranti was forced to “punch out” from LLTV 1 when large lateral-control oscillation developed as he descended from a maximum altitude of 550 feet. Ejecting at 200 feet, Algranti, who had flown the LLTV more than thirty times, landed by parachute uninjured, while the $1.8-million vehicle crashed and burned several hundred feet away. Once again, Houston convened an accident investigation board, headed this time by astronaut Wally Schirra.
MSC director Bob Gilruth and MSC’s Director of Flight Operations Chris Kraft both felt that it was only a matter of time before an astronaut would be killed in the blasted instrument. “Gilruth and I were ready to eliminate it completely,” Kraft notes, “but the astronauts were adamant. They wanted the training it offered.”
LLTV flying resumed in April 1969, even before Schirra’s accident investigation panel turned in its report. When nothing went wrong in the first few flights involving only MSC test pilots, routine training flights for the astronauts began again. Yet Kraft knew they were pressing their luck. One day in the late spring of 1969, he asked Armstrong to stop by his office, hoping that Neil would give him some tidbit that Kraft could turn into a negative report about the LLTV. Neil did not provide it. “It’s absolutely essential,” the commander of the upcoming Apollo 11 mission told him. “By far the best training for landing on the Moon.” “It’s dangerous, damn it!” Kraft snapped back. “Yes, it is,” said Neil. “I know you’re worried, but I have to support it. It’s just darned good training.”
Kraft gave in, but he didn’t give up. Even after the lunar landings began, either he or Gilruth “grilled every returning astronaut, hoping to find some way to get the LLTV grounded forever.” They lost every time, because the astronauts wanted it. “To a man,” Kraft recalls, “they said it was the best training they received and was essential to landing on the Moon. So with our fingers crossed, we let them keep it.” The last astronaut to fly it was Gene Cernan on November 13, 1972, three weeks before the launch of Apollo 17, the final landing mission.
For three straight days in mid-June 1969, less than a month before the launch of Apollo 11, Armstrong flew one of the new LLTVs while Kraft and other NASA managers held their breath. Over the course of those three days (the fourteenth through the sixteenth), he took the LLTV up for lunar descents eight separate times. In all, he made a grand total of nineteen flights in the converted LLRVs and eight flights in the new LLTVs. No other astronaut before or after Armstrong flew the vehicle so much; for the record, Aldrin never flew the machine.19 To observing newsmen after one of his flights, Neil remarked, “We are very pleased with the way it flies. I think it does an excellent job of capturing the handling characteristics. We’re getting a very high level of confidence in the overall landing maneuver.”
To this day, Armstrong remains convinced that the LLTV was “absolutely required to prepare yourself properly for the lunar landing.” If he had not felt this way back in 1968 and 1969, given how contrary and risky he knew the machine was, he would have said, in the best interest of the overall program, “let’s quit it.” Considering his collaboration with his colleagues at Edwards in conceptualizing and developing the original LLRV, “it is human nature to defend the things that you’ve been involved in the creation of. But I really believe that my motivation to recommend using it was always proper.”
Gene Matranga puts the capstone on the significance of the landing simulator for the pilot of the first Moon landing: “Psychologically, it didn’t hurt that the LLRV/LLTV was harder to fly than the LM. That pleasant surprise had to bolster any astronaut’s confidence on his way down to the lunar surface.”