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 gr
ound 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 (but did not occur until Apollo 9 in March 1969). 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.”
First Man Page 40