Category X-15 EXTENDING THE FRONTIERS OF FLIGHT

The public law that established the NACA required the agency to disseminate information to the industry and the public. One of the methods used to accomplish this was to hold periodic conferences with representatives of the industry to discuss the results of research into specific areas. By the beginning of July, Hugh Dryden concluded there had been sufficient progress on the development of the X-15 to hold an industry conference at one of the NACA facilities in October.-411

Langley hosted the first Conference on the Progress of the X-15 Project on 25-26 October 1956, providing an interesting insight into the X-15 development effort. There were 313 attendees representing the Air Force, NACA, Navy, various universities and colleges, and most of the major aerospace contractors. Approximately 10% of the attendees were from various Air Force organizations, with the WADC contributing over half. Oddly, however, Air Force personnel made none of the presentations at the conference. The majority of the 27 authors of the 18 technical papers came from various NACA organizations (16), while the rest were from North American (9) and Reaction Motors (2). The papers confirmed a considerable amount of progress, but made it clear that a few significant problems still lay ahead.-42

Another paper summarized the results of tests in eight different wind tunnels. These tests were conducted at velocities between low subsonic speeds to Mach 6.9, somewhat in excess of the projected maximum speed of the airplane. One of the surprising findings was that the controversial fuselage tunnels generated nearly half of the total lift at high Mach numbers. However, another result confirmed the NACA prediction that the original fuselage tunnels would cause longitudinal instability. In subsequent testing, researchers shortened the tunnels ahead of the wing, greatly reducing the problem.-43

One of the more interesting experiments was "flying" small (3- to 4-inch) models in the hypervelocity free-flight facility at Ames. The models, which were made of cast aluminum, cast bronze, or various plastics, were fragile. Despite this, the goal was to shoot the model out of a gun at tremendous speeds in order to observe shock-wave patterns across the shape. As often as not, what researchers saw were pieces of X-15 models flying down the range sideways. Fortunately, enough of the models remained intact for them to acquire meaningful data.-444

The hypervelocity free-flight facility at NACA Ames fired small (3-4-inch-long) models of the X – 15 to observe shock-wave patterns. It was more of an art than a science to get the models to fly forward and not break apart, but enough survived to gain significant insight into shock patterns surrounding the X-15. (NASA)

Other papers dealt with the ability of the pilot to fly the airplane. Pilots had flown the preliminary exit and reentry profiles using fixed-base simulators at Langley and North American. Alarmingly, the pilots found that the airplane was nearly uncontrollable without damping and only marginally stable during some maneuvers with dampers. A free-flying model program at the PARD showed that low-speed stability and control were adequate. Since some aerodynamicists had questioned the use of the rolling tail instead of ailerons, free-flying models had investigated that feature, proving that the rolling tail would provide the necessary lateral control.[45]

Researchers also reported on the state of the structural design. Preliminary estimates showed that the airplane would encounter critical loads during the initial acceleration and during reentry, but would experience maximum temperatures only during the latter. Because of this, the paper primarily dealt with the load-temperature relationships anticipated for reentry. The selection of Inconel X was justified based on its strength and favorable creep characteristics at 1,200°F. The leading edge would use a bar of Inconel X, since that portion of the wing acted as a heat sink. This represented a radical change from the fiberglass leading edge originally proposed by North American. In another major change, the leading edge of the wing was no longer easily removable, although this fact seemed to escape the attention of most everybody in attendance, particularly Harry Goett from Ames.[46]

The main landing gear brought its own concerns. Originally, it consisted of two narrow skids attached to the fuselage under the front part of the wing and stowed externally along the side tunnels during flight. When unlocked, the skis fell into the down position, with help from airflow and a bungee. Further analysis indicated that the X-15 would land more nose-high than expected, and that the rear fuselage would likely strike the ground before the skids. A small tail – skid had been proposed, but this was found to be inadequate. In its place, engineers moved the skids aft to approximately the leading edge of the vertical stabilizers, solving the ground-strike problem. However, the move introduced a new concern. Now the nose-down rotation after main – skid contact would be particularly jarring, placing a great deal of stress on the pilot and airframe. In fact, it would lead directly to one early landing accident and be a source of problems throughout the flight program. Nobody had a suitable solution.[47]

The expected acceleration of the X-15 presented several unique human-factor concerns early in the program. It was estimated that the pilot would be subjected to an acceleration of up to 5 g. Because of this, North American developed a side-stick controller that used an armrest to support the pilot’s arm while still allowing full control of the airplane. Coupled with the fact that there were two separate attitude-control systems on the X-15, this resulted in a unique control-stick arrangement. A conventional center stick, similar to that installed in most fighter-type aircraft of the era, operated the aerodynamic control surfaces through the newly required stability augmentation (damper) system. Mechanical linkages connected a side-stick controller on the right console to the same aerodynamic control surfaces and augmentation system. The pilot could use either stick interchangeably, although the flight manual described the use of the center stick "during normal periods of longitudinal and vertical acceleration." Another side-stick controller above the left console operated the ballistic control system that provided attitude control at high altitudes. Describing one of the phenomena soon to be discovered in space flight, the flight manual warned that "velocity tends to sustain itself after the stick is returned to the neutral position. A subsequent stick movement opposite to the initial one is required to cancel the original attitude change." Isaac Newton was correct after all.[48]

From the left, North American test pilot Alvin S. White, Air Force X-15 Project Pilot Captain Iven C. Kinchloe, and Scott Cross field discuss the design of the side stick controller for the new research airplane. The design of these controllers caused quite a bit of controversy early in the program, but the pilots generally liked them once they acclimated. Crossfield’s influence on the program showed early in the flight program when some pilots complained the configuration of the cockpit was tailored to Crossfield’s size and was not sufficiently adjustable to accommodate other pilots. Later modifications solved these issues. (Alvin S. White Collection)

Engineers had not firmly established the design for the X-15 side-stick controller, but researchers discussed previous experience with similar controllers in the Convair F-102, Grumman F9F, Lockheed TV-2, and North American YF-107A, as well as several ground simulators. The pilots who had used these controllers generally thought that the engineers needed to provide a more "natural" feel for the controllers.-49

Based largely on urgings from Scott Crossfield, the Air Force agreed to allow North American to use an ejection seat instead of a capsule system. The company had investigated four escape systems in depth, including cockpit capsules, nose capsules, a canopy-shielded seat, and a stable-seat with a pressure suit. Engineers had tried capsule-like systems before, most notably in the X-2, where the entire forward fuselage could be detached from the rest of the aircraft.

Douglas had opted for this approach in all of the D-558s and their X-15 proposal. Model tests showed that these were unstable and prone to tumble at a high rate of rotation, and they added weight and complexity to the aircraft. Their potential success rate was unknown at the time.-501

North American performed a seemingly endless series of analyses to support their selection of an ejection seat over an encapsulated system. The company determined there was only a 2-percent likelihood of an accident occurring at high altitude or high speed, eliminating much of the perceived need for the complicated and heavy encapsulated system. The stabilized ejection seat, coupled with the David Clark Company full-pressure suit, provided meaningful ejection up to Mach 4 and 120,000 feet. (North American Aviation)

Surprisingly, an analysis by North American showed that only 2% of the accidents would occur at high altitude or speed. Because engineers expected most potential accidents to occur at speeds less than Mach 4, North American had decided to use a stable-seat with a pressure suit. The perceived benefits of this combination were its relative simplicity, high reliability, and light weight. North American acknowledged that the seat did not provide meaningful escape at altitudes above 120,000 feet or speeds in excess of Mach 4. However, the designers (particularly Scott Crossfield) believed that when the seat-suit combination was inadequate, the safest course of action was for the pilot to simply ride the airplane down to an altitude and velocity where the ejection seat could function successfully.-1511

Lawrence P. Greene, the chief of aerodynamics at North American, presented the final paper at the 1956 industry conference. This was an excellent summary of the development effort to date and a review of the major known problems. Researchers considered flutter to be a potential problem, largely because little experimental data regarding flutter at hypersonic Mach numbers were available, and there was a lack of basic knowledge on aero-thermal-elastic relationships. Greene pointed out that engineers had derived the available data on high-speed flutter from experiments conducted at less than Mach 3, and not all of it was applicable to the X-15. As it turned out, the program did encounter panel flutter during the early flights, leading to a change in the design criteria for high-speed aircraft.-152

Inconel X also presented a potential problem because fabrication techniques for large structures did not exist. By using various alloys of titanium, North American saved considerable weight in parts of the internal structure that were not subject to high temperatures. Titanium, while usable to only about 800°F, weighed much less than Inconel X. Ultimately, the requirements for processing and fabricating these materials influenced some aspects of the structural design. Inconel X soon stopped being a laboratory curiosity as the X-15 program developed techniques to form, machine, and heat-treat it.[53]

Overall, the conference was a success and disseminated a great deal of information to the industry, along with frank discussions about unresolved issues and concerns. It also provided a short break for the development team that had been working hard to meet an extremely ambitious schedule.

The XLR99 presented several unique challenges to Reaction Motors. Perhaps the major one was that the engine was being developed for a manned vehicle, which entailed more safety and reliability requirements than unmanned missiles. However, perhaps even more challenging were the requirements to be able to throttle and restart the engine in flight-something that had not yet been attempted with a large rocket engine. The Reaction Motor representative at the 1956 industry conference concluded his presentation with the observation that developing the XLR99 was going to be challenging. Subsequent events proved this correct.-138

Robert W. Seaman from Reaction Motors presented preliminary specifications for the XLR99-RM-1 at the conference. The oxygen-ammonia engine could vary its thrust from 19,200 lbf (34%) to 57,200 lbf at 40,000 feet, and had a specific impulse between 256 seconds and 276 seconds depending on the altitude and throttle setting. The engine fit into a space 71.7 inches long and 43.2 inches in diameter. At this point, Reaction Motors was predicting a 618-pound dry weight and a 748-pound gross weight. A two-stage impulse turbine drove the single-inlet oxidizer pump and two-inlet fuel pump. The hydrogen-peroxide-driven turbopump exhausted into the thrust chamber. Regulating the amount of hydrogen peroxide that was decomposed to drive the turbopump provided the throttle control.-139

LOX

i—cm

lOX ‘

ACCUMULATOR

GENERATOR

BASIC ENGINE SCHEMATIC

Although not the most powerful rocket engine of its era, the XLR99 was the most advanced and used a sophisticated turbopump to supply liquid oxygen and anhydrous ammonia propellants to the combustion chamber. The engine was capable of being restarted in flight, an unusual feature for the time (or even today) and numerous safety systems automatically shut down the engine in the event of a problem. (NASA)

Engineers decided to control thrust by regulating the speed of the turbopump because the other possibilities resulted in the turbopump speeding up as pressure decreased, resulting in cavitation. Controlling the propellant to the turbopump also required fewer controls and less instrumentation. However, varying the fuel flow led to other issues, such as how to provide adequate coolant (fuel) to the thrust chamber.[40]

The engineers also had to give engine compartment temperatures more consideration than they did for previous engines due to the high heat transfer expected from the X-15 hot-structure. This was one of the first instances in which the surrounding airframe structure would be hotter than the engine. Since North American was designing the hot structure of the X-15 to withstand temperatures well in excess of those the engine produced, the engineers were not planning to insulate the engine compartment.-41

Another paper discussed engine controls and instruments, accessory installation, and various propellant system components. The 1,000-gallon liquid-oxygen tank was located just ahead of the aircraft center of gravity, and the 1,400-gallon anhydrous-ammonia tank was just behind it. A 3,600-psi helium supply tube within the liquid-oxygen tank supplied the gas to pressurize both tanks. A 75-gallon hydrogen-peroxide tank behind the ammonia tank provided the monopropellant for the turbopump, using a small, additional supply of helium.-421

The liquid-oxygen and ammonia tanks had triple compartments arranged to force the propellants toward the center of gravity during normal operations and during jettisoning. The design needed to compensate for the acceleration of the X-15, which tended to force propellants toward one end of the tanks or the other. Further complicating the design of the tanks was the necessity for efficient loading and minimizing the remaining propellant after burnout or jettisoning.

Fortunately, the tanks did not present any insurmountable problems during early tests.-431

Because the engineers did not yet fully understand the vibration characteristics of the XLR99, they designed a rigid engine mount without any special vibration attenuation. The engine-mount truss attached to the fuselage at three fittings, and by adjusting the lower two fittings the engineers could tailor the thrust vector of the engine. Three large removable doors in the aft fuselage provided access to the engine and allowed closed-circuit television cameras to observe the engine during ground testing. Ultimately, this mounting technique would also make it much easier to use the interim XLR11 engines.-441

The Navy, an otherwise silent partner, made a notable contribution to flight simulation for the X – 15 program. Primarily, the Aviation Medical Acceleration Laboratory (AMAL) at NADC Johnsville provided a unique ground simulation of the dynamic environment.-1281

Even prior to the beginning of World War II, researchers recognized that acceleration effects experienced during high-speed flight would require evaluation, and by 1944 the BuAer became convinced that it would require a long-term commitment to understand such effects completely. The centerpiece of what became the AMAL was a new $2,381,000 human centrifuge. Work on the facility at Johnsville began in June 1947, with the McKiernan-Terry Corporation of Harrison, New Jersey, constructing the centrifuge building under the direction of the Office of Naval Research.

The chief of naval operations established the AMAL on 24 May 1949, and during validation of the facility on 2 November 1951, Captain J. R. Poppin, the director of AMAL, became the first human to be tested in the centrifuge.-291

When the facility officially opened on 17 June 1952, it was the most sophisticated of its kind in the world, and was capable of producing accelerations up to 40 g to investigate the reaction of pilots to accelerations. A 4,000-horsepower vertical electric motor in the center of the room drove the centrifuge arm. Depending on the exact requirements of the test, researchers could position a gondola suspended by a double gimbal system at one of several locations along the arm. The outer gimbal permitted rotation of the gondola about an axis tangential to the motion of the centrifuge, while the inner gimbal allowed rotation about the axis at right angles to the tangential motion. Separate 75-horsepower motors connected through hydraulic actuators controlled the angular motions of the gondola, and continuous control of the two axes in combination with rotation of the arm produced somewhat realistic high-g accelerations for the pilot.201

Initially, electromechanical systems controlled the centrifuge since general-purpose computers did not, for all intents, yet exist. In the centrifuge, large Masonite discs called "cams" controlled the acceleration along the three axes. A series of cam followers drove potentiometers that generated voltages to control the various hydraulic actuators and electric motors. The cams had some distinct advantages over manual control: they automated complex motions and allowed precise duplication of the motions. However, the process of cutting the Masonite discs amounted to little more than trial and error, and technicians had to produce many discs for each test.211

Researchers demonstrated the capabilities of the centrifuge in a series of experiments, including a joint Navy-Air Force study during 1956 that revealed that chimpanzees were able to sustain 40 g for 60 seconds. Two years later R. Flanagan Gray of the NADC set a human record of 31.25 g, which he sustained for 5 seconds in the "iron maiden," a water-filled protective apparatus attached 40 feet out on the arm. In 1957 the X-15 program became the first user of the combined human centrifuge and NADC computer facility, marking the initial step in the development of dynamic flight simulation.-1321

The X-15 represented the most extensive, and by far the most elaborate, use of the cams for centrifuge control. Technicians at Johnsville cut the cams based on acceleration parameters defined by researchers at North American. Initially, the tests concentrated on routine flights, measuring the pilot’s reactions to the accelerations. Before long, the tests were expanded to emergency conditions, such as an X-15 returning from a high-altitude mission with a failed pitch damper. The concern was whether the pilot could tolerate the accelerations expected under these conditions, which included oscillations between 0 g and 8 g on a cycle of 0.7 seconds. Other conditions included oscillations between 4 g and 8 g with periods as long as 12 seconds. Researchers found that these conditions represented something near the physiological tolerance of the pilots. Even with the best support apparatus the engineers could provide, the pilots found it difficult to operate the controls, and small, purplish hemorrhages known as petechiae would form on their hands, feet, and back. In one experiment, Scott Crossfield actually blacked out due to a malfunction in his g-suit.-1331

When NADCJohnsville officially opened on 17 June 1952 it was the most sophisticated human centrifuge in the world, capable of producing accelerations up to 40 g to investigate the reaction of pilots to accelerations. The initial runs at Johnsville used a generic cockpit that did not resemble an X-15 at all. During an early series of tests, researchers mounted an oscilloscope in front of the pilot, and asked him to move the gondola to match a trace on the scope. For the first runs, the pilot used a conventional center stick; later tests used a side-stick controller. (U. S. Navy)

With the use of the Masonite disc cam followers, the gondola was able to maintain a programmed and precisely reproducible acceleration pattern. This was a flaw in some people’s minds since the pilot did not influence the motion of the gondola-he was, in effect, a passenger. However, the X – 15 pilot had to maintain precise control while being forced backward or forward under the high accelerations, and it was important to find out how well he could perform. This was especially true during marginal conditions, such as a damper failure during reentry. There were no guidelines for defining the degree of control expected from a pilot under those conditions.-1341

To address this issue, researchers subsequently modified the centrifuge to incorporate responses to pilot input into the preprogrammed acceleration curves. During an early series of tests, researchers mounted an oscilloscope in front of the pilot and asked him to move the gondola to match a trace on the scope. For the first runs the pilot used a conventional center stick; later tests used a side-stick controller. Eventually the complexity of the acceleration patterns moved beyond the capabilities of the Masonite discs and researchers began using punched paper tape, something that found widespread use on early computers. The results of these experiments indicated that under extreme conditions the side-stick controller allowed the pilot to brace his arm against the cockpit side console to maintain better control of the aircraft.-1351

Researchers at Johnsville soon installed a complete X-15 instrument panel in the gondola, with the instruments receiving data from analog computers to emulate the flight profile being "flown" by the centrifuge. These simulations led to a recommendation to rearrange some of the X-15 instruments to reduce eye movement. As acceleration increased, the pilot’s field of view became narrower, and under grayout conditions the pilots could not adequately scan instruments that were normally in their field of view. Moving a few instruments closer together allowed the pilot to concentrate on one area of the instrument panel without having to move his head, an often difficult and occasionally impossible task under heavy g-loading.[36]

Another important conclusion drawn from this set of experiments was that the centrifuge was sufficiently flexible to use as a dynamic flight simulator. To enable this, in June 1957 researchers linked the centrifuge to the Typhoon analog computer, which was generally similar to the units used in the X-15 fixed-base simulators. This made dynamic control possible, and pilots in the centrifuge gondola could actually "fly" the device, simulating the flight characteristics of any selected type of aircraft. The computer output drove the centrifuge in such a manner that the pilot experienced an approximation of the linear acceleration he would feel while flying the X-15 if he made the same control motions. Unfortunately, the centrifuge only had three degrees of freedom (one in the main arm and two in the gondola gimbal system), whereas the X-15 had six degrees of freedom (three of rotation and three of translation). This meant that the angular accelerations were unlike those experienced in flight; however, researchers believed this limitation was of secondary importance. The perceived benefit of simulating even somewhat unrealistic movements was that they could introduce the pilot to the large accelerations he would experience during flight. The computer also drove the cockpit instruments to reflect the "reality" of flight. Engineers had not previously attempted this type of closed-loop simulation (pilot to computer to centrifuge), and it was a far more complex problem than developing the fixed-base simulators. Interestingly, in an experiment that was years ahead of its time, researchers using the X-15 simulation computer at NASA Langley controlled the Johnsville centrifuge over a telephone line on several occasions. The response time from this arrangement was less than ideal because of the low data rates possible at the time, but the overall concept worked surprisingly well.[37]

Certain inadequacies in the X-15 simulation were noted during these initial tests, particularly concerning the computation of aircraft responses at high frequencies, the pilot restraints, and the lack of simulated speed brakes. In May 1958 the Navy modified the centrifuge in an attempt to cure these problems, and researchers completed three additional weeks of X-15 tests on 12 July 1958. During this time the pilots (Neil Armstrong, Scott Crossfield, Iven Kincheloe, Jack McKay, Joe Walker, Al White, and Bob White) and various other personnel, such as Dick Day and Bob Hoey, flew 755 static simulations using the cockpit installed in the gondola but with the centrifuge turned off. The pilots also completed 287 dynamic simulations with the centrifuge in motion. The primary objective of the program was to assess the pilot’s ability to make emergency reentries under high dynamic conditions following a damper failure. The results were generally encouraging, although the accelerations were more severe than those experienced later during actual flight.138

A typical centrifuge run for a high-altitude mission commenced after the pilot attained the exit flight path and a speed of Mach 2, and terminated after the pilot brought the aircraft back to level flight after reentry. During powered flight, the thrust acceleration gradually built up to 4.5 g, forcing the pilot against the seat back. However, the pilot could keep his feet on the rudder pedals with some effort, and still reach the instrument panel to operate switches if required. Researchers also simulated the consequences of thrust misalignment so that during powered flight the pilot would know to apply aerodynamic control corrections with the right-hand side stick and the rudder pedals.-139

At burnout, the acceleration component dropped to zero and the pilot’s head came off the backrest. The pilot attempted to hold the aircraft heading using the ballistic control system. In the design mission, the aircraft would experience less than 0.1 g for about 150 seconds, but the best the centrifuge could do was to remain at rest (and 1 g) during this period since there was no way to simulate less than normal gravity.-139

A 4,000-horsepower vertical electric motor in the center of the room drove the centrifuge arm that had a gondola suspended by a double gimbal system at one of several locations along the arm. The outer gimbal permitted rotation of the gondola about an axis tangential to the motion of the centrifuge; the inner gimbal allowed rotation about the axis at right angles to the tangential motion. Continuous control of the two axes in combination with rotation of the arm produced somewhat realistic high-g accelerations for the pilot in the gondola. Johnsville would gain fame when the Mercury program used the centrifuge for much the same purposes the X-15 had pioneered several years earlier. (U. S. Navy)

As the aircraft descended, the pilot actuated the pitch trim knob and the aerodynamic control stick at about 200,000 feet to establish the desired angle of attack, but continued to use the ballistic control system until the aerodynamic controls became effective. As the dynamic pressure built, the pullout acceleration commenced and the centrifuge began to turn. If the speed brakes were closed, the drag deceleration reached about 1 g. With the speed brakes open, this would increase to 2.8 g for the design mission and about 4 g for a reentry from 550,000 feet. The pilot gradually reduced the angle of attack to maintain the designed g-value until the aircraft was level, at which time the simulation stopped. During reentry, in addition to the drag acceleration, the pilot also experienced 5-7 g of normal acceleration, so the total g-vector was 6-8 g "eyeballs down and forward"-a very undesirable physiological condition.*41

Tests on the centrifuge established that, with proper restraints and anti-g equipment, the pilot of the X-15 could tolerate the expected accelerations. These included such oscillating accelerations as 5 g 2 g at one cycle per second for 10 seconds, which might occur during reentry from 250,000 feet with failed dampers, and 7 g normal and 4 g "into the straps" for 25 seconds, which might occur during reentry from 550,000 feet. The pilots’ ability to tolerate oscillating accelerations was unknown prior to the centrifuge tests, and this information contributed not only to the X-15 but also to Mercury and later space programs.*421

The tests at Johnsville confirmed that a trained pilot could not only tolerate the acceleration levels, he could also perform all tasks reasonably expected of him under those conditions. This was largely due to the North American design of pilot supports and restraints, and the use of side- stick controllers. The accommodations included a bucket seat without padding adjusted in height for each pilot, and arm and elbow rests also fitted for each pilot. Restraints included an integrated harness with the lower ties lateral to the hips to minimize "submarining" and rolling in the seat, a helmet "socket" to limit motion posteriorly, laterally, and at the top, and a retractable front "head bumper" that could be swung down to limit forward motion of the head. When using the speed brakes or when the dampers were off, the pilots generally found it desirable to use the front head bumper. The pilots used the centrifuge program to evaluate two kinematic designs and three grip designs for the side-stick controller before an acceptable one was found. Despite an early reluctance, the pilots generally preferred the side stick to the center stick under dynamic conditions. Researchers quickly established the importance of careful dynamic balancing and suitable breakout and friction forces for the side stick.*431

The centrifuge program also pointed out the need for pilot experience under high-acceleration conditions. For example, pilots who had at least 15 hours of practice on the static simulator at Inglewood and previous high-acceleration experience made five successful dynamic reentries out of five attempts, while pilots with 4-10 hours of simulator time had only seven successes in 15 attempts. Another group of pilots who had less than 4 hours of simulator time or no previous high-acceleration experience made only two successful dynamic reentries out of 14 attempts.

Most of the failures were due to unintentional pilot control inputs, including using the rudder pedals during drag deceleration, roll inputs while making pitch corrections using the center stick because of the lack of arm support, and inadvertent ballistic control system firings due to leaving the left hand on the side-stick during acceleration. The more experienced pilots would detect these unintended control inputs more rapidly than the other pilots, and could correct the mistakes in time to avoid serious consequences.*441

Researchers also evaluated physiological responses in the centrifuge. The drag decelerations of the speed brakes, when combined with the normal pullout loads, increased the blood pressure in the limbs. When the resultant acceleration was below 5 g, there was no particular discomfort; however, when the acceleration was above 7 g (including a drag component of more than 3 g), petechiae were noted in the forearms and ankles, and a tingling, numbness, and in some cases definite pain were noted in the limbs. The symptoms became more severe when a pilot made several centrifuge runs in quick succession, something that would obviously never happen during the X-15 program. One pilot stopped the centrifuge when he experienced severe groin pains because of a poorly fitted harness. In two cases of reentry using open speed brakes, the pilots reported pronounced oculogravic illusions, with the visual field seeming to oscillate vertically and to be doubled vertically for a few seconds toward the end of the reentry. Despite this, Scott Crossfield made nine dynamic runs in one day on the centrifuge, but generally the pilots were limited to two runs on the centrifuge per day.[45]

Despite the demonstrated benefits of a pilot being able to experience the unusually high accelerations produced by the X-15 prior to his first flight, only the initial group of pilots actually benefited from the centrifuge simulations. Later pilots received the surprise of their life the first time they started the XLR99 in the X-15. Granted, the Johnsville accelerations were not a realistic replica of the ones experienced in flight, due to the limitations of the centrifuge concept, but they still provided some high-acceleration experience. As Milt Thompson noted in a paper in 1964:[46]

Prior to my first flight, my practice had been done in a relaxed, head forward position. The longitudinal acceleration at engine light forced my head back into the headrest and prevented even helmet rotation. The instrument-scan procedure, due to this head position and a slight tunnel vision effect, was quite different than anticipated and practiced. The acceleration buildup during engine burn (4-g max) is uncomfortable enough to convince you to shut down the engine as planned. This is the first airplane I’ve flown that I was happy to shut down. Engine shutdown does not relieve the situation, though, since in most cases the deceleration immediately after shutdown has you hanging from the restraint harness, and in a strange position for controlling [the airplane].

The X-15 closed-loop program was the forerunner of centrifuges that NASA built at the Ames Research Center and the Manned Spacecraft Center (later renamed the Johnson Space Center) to support the manned space programs. Perhaps the most celebrated program of AMAL was the flight simulation training for Project Mercury astronauts, based largely on the experience gained during the X-15 simulations. Beginning in June 1959 the seven Mercury astronauts participated in centrifuge simulations of Atlas booster launches, reentries, and abort conditions ranging up to 18 g (transverse) at NADC Johnsville.[47]

Not surprisingly, during the early 1950s the top priority for the hypersonic tunnels was to support the massive development effort associated with the intercontinental missiles then under development. Initially it was not clear whether the resulting weapon would be a high-speed cruise missile or an intercontinental ballistic missile (ICBM), so the Air Force undertook programs to develop both. Much of the theoretical science necessary to create a manned hypersonic research airplane would be born of the perceived need to build these weapons. Long-range missile development challenged NACA researchers in a number of ways. The advancements necessary to allow a Mach 3 cruise missile were relatively easily imagined, if not readily at hand. The ballistic missile was a different story. A successful ICBM would have to accelerate to 15,000 miles per hour at an altitude of perhaps 500 miles, and then be guided to a precise target thousands of miles away. Sophisticated and reliable propulsion, control, and guidance systems were essential, as was keeping the structural weight at a minimum. Moreover, researchers needed to find some method to handle aerodynamic heating. As the missile warhead reentered the atmosphere, it would experience temperatures of several thousand °F. The heat that was generated by shock-wave compression outside the boundary layer and was not in contact with the structure would dissipate harmlessly into the surrounding air. However, the part that arose within the boundary layer and was in direct contact with the missile structure would be great enough to melt the vehicle. Many early dummy warheads burned up because the engineers did not yet understand this.

During this time, H. Julian Allen was engaged in high-speed research at Ames and found what he believed to be a practical solution to the aerodynamic heating problems of the ICBM. In place of the traditional sleek configuration with a sharply pointed nose (an aerodynamic concept long since embraced by missile designers, mostly because the V-2 had used it), Allen proposed a blunt shape with a rounded bottom. In 1951 Allen predicted that when the missile reentered the atmosphere, its blunt shape would create a powerful bow-shaped shock wave that would deflect heat safely outward and away from the structure of the missile. The boundary layer on the body created some frictional drag and heating, but this was only a small fraction of the total heat of deceleration, most of which harmlessly heated the atmosphere through the action of the strong shock wave. As Allen and Eggers put it, "not only should pointed bodies be avoided, but the rounded nose should have as large a radius as possible." Thus the "blunt-body" concept was born.[52]

In 1951, NACA Ames researcher H. Julian Allen postulated the concept of a "blunt body" reentry vehicle for intercontinental missiles. Pushing the shock wave away from the missile body removed most of the aerodynamic heating from being in direct contact with the structure. The reentry profiles developed at NASA Langley used the idea of "sufficient lift," which were a new manifestation of the blunt-body concept. (NASA)

Allen and Eggers verified the blunt-body concept by studying the aerodynamic heating of miniature missiles in an innovative supersonic free-flight tunnel, a sort of wind-tunnel-cum – firing-range that had become operational at Ames in 1949. The researchers published their classified report on these tests in August 1953, but the Air Force and aerospace industry did not immediately embrace the concept since it ran contrary to most established ideas. Engineers accustomed to pointed-body missiles remained skeptical of the blunt-body concept until the mid-to-late-1950s, when it became the basis for the new ICBM warheads and all of the manned space capsules.-153

In the meantime, Robert J. Woods, designer of the Bell X-1 and X-2 research airplanes, stirred up interest in hypersonic aircraft. In a letter to the NACA Committee on Aerodynamics^ dated 8 January 1952, Woods proposed that the committee direct some part of its research to address the basic problems of hypersonic and space flight. Accompanying the letter was a document from Dr.

Walter R. Dornberger, former commander of the German rocket test facility at Peenemunde and now a Bell employee, outlining the preliminary requirements of a hypersonic aircraft. The "ionosphere research plane" proposed by Dornberger was powered by a liquid-fueled rocket engine and capable of flying at 6,000 feet per second (fps) at an altitude of 50-75 miles.-1551 It was apparent that the concept for an "antipodal" bomber proposed near the end of the war by his colleagues Eugen Sanger and Irene Bredt still intrigued Dornberger.-1551 According to the Sanger – Bredt study, this aircraft would skip in and out of the atmosphere (called "skip-gliding") and land halfway around the world.1571 Dornberger’s enthusiasm for the concept had captured Woods’s imagination, and he called for the NACA to develop a manned hypersonic research airplane in support of it. At the time, the committee declined to initiate the research advocated by Woods, but took the matter under advisement.1581

At the 30 January 1952 meeting of the Committee on Aerodynamics, Woods submitted a paper that noted growing interest in very-high-speed flight at altitudes where the atmospheric density was so low as to eliminate effective aerodynamic control. Since he believed that research into this regime was necessary, Woods suggested that "the NACA is the logical organization to carry out the basic studies in space flight control and stability" and that the NACA should set up a small group "to evaluate and analyze the basic problems of space flight." Woods went on to recommend that the NACA "endeavor to establish a concept of a suitable manned test vehicle" that could be developed within two years. Again, the NACA took the matter under advisement.1591

Smith J. DeFrance, an early Langley engineer who became the director of NACA Ames when it opened in 1941, opposed the idea for a hypersonic study group because "it appears to verge on the developmental, and there is a question as to its importance. There are many more pressing and more realistic problems to be met and solved in the next ten years." DeFrance concluded in the spring of 1952 that "a study group of any size is not warranted." This reflected the position of many NACA researchers who believed the committee should only undertake theoretical and basic research, and leave development projects to the military and industry.1601

Further discussion ensued during the 24 June 1952 meeting of the Committee on Aerodynamics. Other factors covered at the meeting included Allen’s unanticipated discovery of the blunt-body concept and a special request from a group representing 11 missile manufacturers.

The NACA Subcommittee on Stability and Control had invited the same manufacturers to Washington in June 1951 to present their ideas "on the direction in which NACA research should move for greatest benefit in missile development." In this case the weapons in question were more often than not air-to-air and surface-to-air missiles rather than ICBMs. During this meeting, Maxwell W. Hunter, an engineer who was developing the Sparrow and Nike missiles at the Douglas Aircraft Company, suggested that the NACA should begin to explore the problems missiles would encounter at speeds of Mach 4 to Mach 10. Hunter pointed out that several aircraft designers, notably Alexander Kartveli at Republic, were already designing Mach 3 + interceptors.1611 For an air-to-air missile to be effective when launched from an aircraft at Mach 3, the missile itself would most probably need to be capable of hypersonic speeds.1621

Hunter and Woods repeated their requests during the June 1952 meeting of the Committee on Aerodynamics. In response, the committee passed a resolution largely penned by Air Force science advisor Albert Lombard. The resolution recommended that "(1) the NACA increase its program dealing with the problems of unmanned and manned flight in the upper stratosphere at altitudes between 12 and 50 miles, and at Mach numbers between 4 and 10, and (2) the NACA devote a modest effort to problems associated with unmanned and manned flight at altitudes from 50 miles to infinity and at speeds from Mach number 10 to the velocity of escape from

Earth’s gravity." The NACA Executive Committee ratified the resolution on 14 July. NACA Headquarters then asked the Ames, Langley, and Lewis[63] laboratories for comments and recommendations concerning the implementation of this resolution.1641

This resolution had little immediate effect on existing Langley programs, with the exception that it inspired the Pilotless Aircraft Research Division (PARD)-651 to evaluate the possibility of increasing the speeds of their test rockets up to Mach 10. Nevertheless, the resolution did have one very important consequence for the future: the final paragraph called for the laboratories "to devote a modest effort" to the study of space flight.-1661

The concepts and ideas discussed by Dornberger, Hunter, and Woods inspired two unsolicited proposals for research aircraft. The first, released on 21 May 1952, was from Hubert M. "Jake" Drake and L. Robert Carman of the NACA High-Speed Flight Research Station (HSFRS) and called for a two-stage system in which a large supersonic carrier aircraft would launch a smaller, manned research airplane. The Drake-Carman proposal stated that by "using presently available components and manufacturing techniques, an aircraft having a gross weight of 100,000 pounds could be built with an empty weight of 26,900 pounds. Using liquid oxygen and water-alcohol propellants, this aircraft would be capable of attaining Mach numbers of 6.4 and altitudes up to 660,000 feet. It would have duration of one minute at a Mach number of 5.3. By using this aircraft, an aircraft of the size and weight of the Bell X-2 could be launched at Mach 3 and an altitude of 150,000 feet, attaining Mach numbers up to almost 10 and an altitude of about 1,000,000 feet. Duration of one minute at a Mach number of 8 would be possible." The report went into a fair amount of detail concerning the carrier aircraft, but surprisingly little toward describing the heating and structural problems expected for the smaller research airplane.-1671

David G. Stone, head of the Stability and Control Branch of the PARD, released the second report in late May 1952. This report was somewhat more conservative and proposed that the Bell X-2 itself could be used to reach speeds approaching Mach 4.5 and altitudes near 300,000 feet if it were equipped with two JPL-4 Sergeant solid-propellant rocket motors. Stone also recommended the formation of a project group that would work out the details of actual hardware development, flight programs, and aircraft systems. Langley director Henry J. E. Reid and John Stack generally supported this approach, but believed that further study of possible alternatives was required.-681

Meanwhile, in response to the 1952 recommendation from the NACA Committee on Aerodynamics, Henry Reid set up a three-man study group consisting of Clinton E. Brown (chairman) from the Compressibility Research Division, William J. O’Sullivan, Jr., from the PARD, and Charles H. Zimmerman from the Stability and Control Division. Curiously, none of the three had any significant background in hypersonics. Floyd L. Thompson, who became associate director of Langley in September 1952, had rejected a suggestion to include a hypersonic aerodynamicist or specialist in thermodynamics in the study group. Thompson’s plan was to bring together creative engineers with "completely fresh, unbiased ideas." The group was to evaluate the state of available technology and suggest possible programs that researchers could initiate in 1954, given adequate funding.-691

This group reviewed the ongoing ICBM-related work at Convair and RAND,-701 and then investigated the feasibility of hypersonic and reentry flight in general terms. Not surprisingly, the group identified structural heating as the single most important problem. The group also reviewed the earlier proposals from Drake-Carman and Stone, and agreed to endorse a version of Stone’s X-2 modification with several changes. In the Langley concept, the vehicle used a more powerful internal rocket engine instead of strap-on solid boosters, with the goal of reaching Mach 3.7 velocities. Dr. John E. Duberg, the chief of the Structural Research Division, noted, however, that "considerable doubt exists about the ability of the X-2 airplane to survive the planned trajectory because of the high thermal stresses." The study group released its report on 23 June 1953, and in a surprisingly conservative vein, agreed that unmanned missiles should conduct any research in excess of Mach 4.5.1741

Originally, the plan was to have an interlaboratory board review the findings of the study group, but this apparently never happened. Nevertheless, hypersonic specialists at Langley frequently had the opportunity to talk with the group, and heard Brown formally summarize the findings at a briefing in late June 1953. While listening to this summary, the specialists "felt a strong sense of deja-vu," especially on hearing Brown’s pronouncement that "the main problem of hypersonic flight is aerodynamic heating." They disagreed, however, with the group’s conclusion that the NACA would have to rely on flight-testing, rather than on ground-based approaches, for research and development beyond Mach 4.[72]

Brown, O’Sullivan, and Zimmerman found it necessary to reject the use of traditional ground facilities for hypersonic research because they were "entirely inadequate" in accounting for the effects of high temperatures.-1731 John Becker later wrote that "much of the work of the new small hypersonic tunnels was viewed with extreme skepticism" because they could not simulate the correct temperatures and boundary-layer conditions. The Brown study anticipated there would be significant differences between the "hot" aerodynamics of hypersonic flight and the "cold" aerodynamics simulated in ground facilities. The study concluded that "testing would have to be done in actual flight where the high-temperature hypersonic environment would be generated" and recommended extending the PARD rocket-model testing technique to much higher speeds. This would also mean longer ranges, and the study suggested it might be possible to recover the test models in the Sahara Desert of northern Africa.-741

This was another case of the free-flight-versus-wind-tunnel debate that had existed at Langley for years. Ground facilities could not simulate the high-temperature environment at very high Mach numbers, admitted the hypersonics specialists, but facilities like the pilot 11-inch hypersonic tunnel at Langley and the 10-by – 14-inch continuous-flow facility at Ames had proven quite capable of performing a "partial simulation." Selective flight-testing of the final article was desirable-just as it always had been—but, for the sake of safety, economy, and the systematic parametric investigation of details, the hypersonics specialists argued that ground-based techniques had to be the primary tools for aerodynamic research. Similar debates existed between the wind-tunnel researchers and the model-rocket researchers at PARD.-1751

Although Langley had not viewed their May 1952 proposal favorably, in August 1953 Drake and Carman wrote a letter to NACA Headquarters calling for a five-phase hypersonic research program that would lead to a winged orbital vehicle. Dr. Hugh L. Dryden, the director of the NACA, and John W. "Gus" Crowley, the associate director for research at NACA Headquarters, shelved the proposal as being too futuristic.1761 Nevertheless, in its bold advocacy of a "piggyback" two-stage – to-orbit research vehicle, the Drake-Carman report presented one of the earliest serious predecessors of the Space Shuttle.

The previous year had resulted in some major configuration changes to the X-15. The wing size and shape were similar to those proposed by North American, but engineers increased the leading-edge radius (along with the radius on the empennage and nose) to satisfy aerodynamic heating concerns. The leading edge was also changed from replaceable fiberglass to a nearly solid piece of Inconel X. NASA had always harbored concerns about the use of ablative materials on the leading edge, but this change also eliminated the removable-leading-edge concept that was highly prized by Ames. The final configuration also increased the diameter of the fuselage by about 6% in order to increase the propellant capacity.-1541

A revised landing gear eliminated tail-strikes during landing and improved directional stability during slide-out. The side fairings, always a point of contention between North American and the NACA, were shortened ahead of the wing. The horizontal stabilizer was moved rearward 5.4 inches, the wing was moved forward 3.6 inches, and the center of gravity was brought forward 10 inches to improve longitudinal stability. However, perhaps the most visible change was that the area of the vertical stabilizers was increased from 50 square feet to 75 square feet. Full 10- degree wedge airfoils replaced the original double-wedge configuration for the vertical stabilizers. The area for the verticals was also redistributed (55% for the dorsal stabilizer and 45% for the ventral, instead of the original 73/27 configuration). In addition, both the dorsal and ventral stabilizers now had rudders that were nearly symmetrical and operated together at all times (except after the ventral had been jettisoned during landing). Originally, only the dorsal stabilizer had a rudder.-155

The development engineering inspection (DEI) took place in Inglewood facility on 12-13 December 1956. In the normal course of development, the Air Force inspected full-scale mockups to ensure the design features were satisfactory before construction of the first airplane began. Of the 49 people who took part in the inspection, 34 were from the Air Force, with the WADC contributing 22. The inspection committee consisted of Major E. C. Freeman from the ARDC, Mr.

F. Orazio of the WADC, and Lieutenant Colonel Keith G. Lindell from Air Force Headquarters. The NACA and the Navy each contributed a single voting member. Captain Chester E. McCollough, Jr., from the X-15 Project Office, Captain Iven C. Kincheloe, Jr. (already selected as the first Air Force X-15 pilot), and three NACA researchers served as technical advisors.-156

The inspection resulted in 84 requests for alterations, of which the board rejected 12 and deferred 22 others for further study. Surprisingly, the board rejected some of the more interesting of the proposed changes. These included suggestions that the aerodynamic center stick should be capable of controlling the ballistic controls at the press of a switch, the motions of the aerodynamic and ballistic side sticks should be similar, or a third controller that combined both functions should be installed on the right console. The committee rejected these suggestions since it seemed inappropriate to make decisions on worthwhile improvements or combinations before evaluating the controllers already selected under actual flight conditions. Given that two of the three controller suggestions came from future X-15 pilots (Iven Kincheloe and Joseph A. Walker), it appeared that improvements were necessary.-157

An even more surprising rejection occurred concerning changeable leading edges. North American had disclosed at the 1956 industry conference six weeks earlier that the leading edges were no longer removable, with little comment. Nevertheless, Harry Goett from Ames did not agree with the change. Goett wanted to widen the front spar lower flange and locate the ballistic roll thrusters at the back of the same spar. In addition, Goett argued that North American had initially proposed providing interchangeable wing leading edges. In spite of these logical arguments, the inspection committee decided the required changes would add 3 pounds to the design and rejected the request. At least one participant opined that deleting this feature would significantly decrease the value of the hypersonic research airplane.-158

The X-15 mockup as it was inspected in December 1956. At this point, the airplane looked substantially as it would in final form with short fuselage tunnels and shorter vertical surfaces. This inspection cleared the way for North American to produce the final manufacturing drawings and begin to cut metal. (U. S. Air Force)

Additional wind-tunnel testing resulted in modifications to the vertical stabilizer, but North American essentially built the configuration inspected in mockup form during December 1956. However, while the design and construction of the airframe progressed relatively smoothly, other systems were running into serious difficulties.

However, North American was becoming concerned about the engine development effort, echoing many of the same concerns expressed by John Sloop at the NACA. At the 1956 industry conference, North American vice president Raymond H. Rice announced that the XLR99 was four months behind schedule.

The Air Force and Reaction Motors held meetings on 12 and 18 February, and the Air Force, the NACA, North American, and Reaction Motors met on 19 February. Data presented at these meetings confirmed that the engine was approximately four months behind schedule and overweight. Although the performance estimates were decreasing, the deterioration appeared to be relatively minor. General Estes wrote Hugh Dryden (and copied Rice) that "every effort will be expended to prevent further engine schedule slippage."*461

The NACA’s reaction to the February meeting was different. Hartley Soule reported that the Air Force accepted the four-month delay, but that Reaction Motors would deliver two engines by 1 September 1958 instead of one. The Air Force also accepted a decrease from 241 to 236 seconds of specific impulse, and a weight increase from 588 to 618 pounds. Soule pointed out that Reaction Motors had not yet conducted any thrust-chamber tests, and expressed doubt that the revised schedule was achievable. He also noted that the Air Force had scheduled additional engine progress meetings for June and September. On the other hand, the NACA agreed to help Reaction Motors optimize the engine nozzle for high-altitude operations in an attempt to recover some performance. Separately, on 29 March 1957 the X-15 Project Office reported that engine costs had increased to an estimated $14,000,000, plus fee.*471

Unfortunately, Hartley Soule’s premonitions proved correct. Reaction Motors informed the Air Force on 10 July 1957 that a nine-month schedule slip would be necessary to meet the February specifications. In addition, the development would cost $21,800,000-a 50% increase in only 100 days. Alternately, for $17,000,000 Reaction Motors could develop a compliant engine within the established schedule if the weight could be increased to 836 pounds from the original 618 pounds. Representatives from the Air Force, the NACA, North American, and Reaction Motors met at Wright Field on 29 July to discuss alternatives. The participants generally considered the performance penalty a lesser concern than the increased cost and schedule slip needed to develop the "specification" engine, and the Air Force elected to pursue the heavier engine. Reaction Motors mitigated some concerns when it subsequently reported that the turbopump was exceeding its performance goals, allowing a 197-pound reduction in hydrogen-peroxide propellant. In effect, this resulted in an engine that was only 51 pounds heavier than the original 588-pound specification.

Unfortunately, serious problems arose during development of the thrust chamber and injector assemblies. Primarily, the oxidizer tubes of the spaghetti-type injector tended to burn through at low thrust levels. The Air Force encouraged the company to redouble its efforts, but agreed to raise the minimum thrust requirement if necessary. The Air Force and Reaction Motors also discussed changing to a spud-type injector, but did not reach a final decision.*481

Despite the increase in weight, the engine program continued to fall behind. On 11 December 1957, during a meeting at the newly formed Propulsion Laboratory, the company reported an additional six-month delay.*491 Reaction Motors attributed this to an explosion that destroyed the first developmental engine, and a series of turbopump failures. The company also confirmed that it had failed to develop a spaghetti-type injector that met the performance and reliability requirements. Overall, the picture was rather bleak.

The spaghetti-type injector consisted of bundled tubing, with each metal tube going to an individual fuel injector. However, Lieutenant K. E. Weiss, the XLR99 project engineer for the Power Plant Laboratory, designed a spud-type injector that used small, perforated disks instead of tubes. Wright Field machine shops built several of the Weiss designs, and researchers ran preliminary tests in early 1958. By March, Reaction Motors was investigating using the spud-type injector on the XLR99.

The mounting engine delays were beginning to threaten the entire X-15 project. In response, WADC commander Major General Stanley T. Wray and Brigadier General Haugen ordered an investigation of the technical and managerial problems. On 7 January 1958, the Air Force asked Reaction Motors to provide a revised schedule and explain how it would correct the various problems. The company submitted the schedule in mid-January, showing a new five-month delay and an increase in costs to $34,400,000-nearly double the July estimate.-1501

Accompanied by personnel from the X-15 Project Office and Propulsion Laboratory, generals Haugen and Wray visited Reaction Motors on 28 January 1958 to discuss the various concerns. Haugen commented on the company’s poor record of accomplishment up to that time, which was especially troubling given the importance of the X-15 project. Reaction Motors admitted to its "past deficiencies" and assured the generals that it could meet the current cost and schedule estimates. Haugen and Wray left only partly convinced.-1511

The Propulsion Laboratory and the X-15 Project Office reported their recommendations to the ARDC and WADC commanders in mid-February, and to the director of research and development in Air Force Headquarters, Major General Ralph P. Swofford, Jr., on 21 February 1958. These recommendations included continuing the Reaction Motors development program, using XLR11 engines for initial X-15 flights, approving overtime, assigning the project a top Department

of Defense priority (DX rating), establishing a Technical Advisory Group, and initiating an alternate engine development program.

Of these recommendations, the Air Force approved the use of XLR11 engines, an increased Reaction Motors effort, additional funds to cover the increased effort, and the establishment of the advisory group. The XLR11 decision hardly came as a surprise to the engineers at the HSFS and Lewis-they had suggested the same thing nearly three years earlier, as had some at Wright Field. Officials at Air Force headquarters denied the request for a top priority, although they approved a slightly improved priority. The X-15 Project Office postponed the decision concerning the development of an alternate engine, and made it clear that there was a clear distinction between proposals for an interim engine for the initial flight tests and an alternate engine to replace the XLR99 in the final X-15.1521

North American had already investigated the idea of installing a pair of XLR11s at the suggestion of L. Robert Carman. Scott Crossfield was not impressed with the idea and said, "I think we’d be making a big mistake." Crossfield was afraid that once the Air Force approved the change, the troublesome larger engine would never be installed, leaving the X-15 a Mach 3+ airplane instead of one twice that fast. Charlie Feltz and Harrison Storms, however, thought the concept had merit. The XLR11 used liquid oxygen, like the XLR99, so the oxidizer tank required no changes. The smaller engine used alcohol instead of ammonia, but the two liquids were roughly comparable and only minor changes were necessary. Feltz, for one, was slightly relieved: "I’ve been a little concerned about busting into space all at once with a brand-new airplane and a brand-new untried engine…. We’re trying to crack space, with a new pressure suit, reentry, new metal, landing—everything at once. I’ve got a real good buddy [Crossfield] who’s going to be flying that airplane for the first time, and I’d just as soon have him around for a while." After a few weeks, even Crossfield came around: "We should learn to crawl before we enter the Olympic hundred – yard dash." Once the government approved the concept of using XLR11s, the technicians at Edwards began assembling a dozen XLR11s from pieces and parts of various XLR11 and LR8 engines left over from previous programs.1531

The recommendations also resulted in the establishment of a Technical Advisory Group consisting of representatives from the ARDC, BuAer, NACA, and WADC. The first meeting was held at the Reaction Motors facility on 24 February 1958, and the group immediately determined that the thrust chamber was the item that could benefit the most from this advice, since it represented the greatest risk.-54

In addition to the Technical Advisory Group, the government enlisted the help of other rocket engineers to develop an alternate thrust chamber. North American, which owned Rocketdyne, was reluctant to become involved given its role as the X-15 airframe contractor. Eventually, however, generals Wray and Haugen convinced Lee Atwood to allow Rocketdyne to assist Reaction Motors and begin development of an alternate thrust chamber and injectors. Once North American overcame its corporate reluctance, Rocketdyne immediately began adapting the thrust chamber and injector from the Atlas ICBM XLR105-NA-1 sustainer engine to the XLR99.[55]

An additional complication soon developed, although it apparently did not significantly affect the development effort; Reaction Motors and the Thiokol Chemical Corporation began merger negotiations in the early part of 1958. During this period the anticipated reorganization undoubtedly created a distracting uncertainty among Reaction Motors management and employees. Reaction Motors Incorporated (RMI) stockholders approved the merger on 17 April 1958, and the company subsequently became the Reaction Motors Division (RMD) of Thiokol Chemical Corporation.-1561

The Air Force decision to bring Rocketdyne into the fray motivated Reaction Motors to consider alternate designs. However, by the end of April the Air Force acknowledged there were not sufficient funds to develop alternate designs from Rocketdyne and Reaction Motors. Believing that the Rocketdyne XLR105 derivative offered the best chance of success, the Powerplant Laboratory urged Reaction Motors to subcontract with Rocketdyne for its development. Reaction Motors evaluated which design offered the most promise and presented the results at a meeting of Reaction Motors, Rocketdyne, NACA, and WADC representatives on 27 May 1958 at Wright Field. The participants concluded that the Reaction Motors concentric shell thrust chamber would not solve the chamber burnout issue, and Reaction Motors did not believe it could complete the design in time to support the flight program in any case. Since this was obviously not acceptable, all parties agreed that Reaction Motors should discontinue its efforts and subcontract with Rocketdyne for the XLR105 derivative. Two days later the Air Force officially transmitted the 27 May decisions to Reaction Motors.-1571

The next day Reaction Motors and Rocketdyne agreed that $500,000 would fund the development effort through mid-July. Rocketdyne estimated it would cost $1,746,756 to develop the alternate thrust chamber. Producing 14 chambers for initial testing would cost $811,244, and 14 flight chambers would add $657,300.[58]

Despite the appearance of progress, neither the Air Force nor the NACA was completely happy with the progress of the engine development effort. The Propulsion Laboratory prepared two letters intended to provide additional motivation for Reaction Motors. The first was from General Wray to General Anderson, dated 17 June 1958:[59]

For some time, General Haugen and I have been concerned by the poor progress made by Reaction Motors Division on the development of the XLR99 rocket engine for the X-15 airplane program. This engine was one that had been recommended…on the strength of a supposed advanced state of development of the LR30 rocket engine.. In spite of this state of development, Reaction Motors Division has experienced continual schedule slippage and financial overruns.. It is by their own admission, as well as the conclusions of our project engineers, that Reaction Motors Division has used poor judgment and management during the early stages of the engine development program. Inability to meet performance and original Preliminary Flight Rating Test initiation date, which was a contractor deficiency, has resulted in submission of supplemental proposals. This by acceptance or rejection has placed the Air Force in the undesirable position of making program decisions which we would have preferred the contractor, through better management, to have made at a much earlier date.

Wray also wrote a second letter addressed to Thiokol president Joseph W. Crosby, but felt it would have more impact if Anderson signed it. Anderson shortened the four-page draft to two pages before he sent it to Crosby on 27 June. Anderson had tempered Wray’s adversarial tone somewhat, but still left little doubt that the Air Force was upset. The letter implied, but never explicitly stated, that cancellation of the entire contract for nonperformance was an option. In retrospect, it was high unlikely that the Air Force would ever have taken such drastic action since it likely would have spelled the end of the X-15 program as well.[60]

It is difficult to determine whether the letters, or even the implied threat to cancel the Reaction Motors contract, had any effect on the program. Regardless, things began to improve. Test engines at Lake Denmark accumulated more firing time during the first two weeks of July than during the entire program to date. The tests showed that performance was somewhat low, but by 7 August 1958, engine performance increased to within 2.5% of the specification. Of course, the "specification" had and would change over the course of the contract, as illustrated below:[61]

Although the maximum thrust remained constant, the decrease in specific impulse along with the increased weight had serious performance implications for the X-15. The change in the minimum thrust had less effect, and greatly simplified the development effort, but even so, the flight program seldom used low throttle settings.

By August it was obvious that Rocketdyne had been rather optimistic. At this point the Reaction Motors subcontract with Rocketdyne had already cost $3,125,000-almost double the original estimate. The Propulsion Laboratory believed this was unreasonable given that the original premise was that the XLR105 was a well-established design that needed only minor changes to adapt it to the XLR99. There had been so little progress that the Propulsion Laboratory suggested the Rocketdyne effort be canceled "as soon as possible."*62*

A meeting held at Reaction Motors on 15 August 1958 included Hartley Soule, Brigadier General Haugen, Brigadier General Waymond A. Davis, and representatives from Air Force Headquarters, the ARDC, and the WADC. Reaction Motors and Rocketdyne provided briefings on the status of their respective efforts, and the participants agreed to freeze the engine design using the Reaction Motors thrust chamber. Reaction Motors was encouraged to continue making minor changes to the injector in an attempt to improve performance, but was cautioned not to delay the schedule or to sacrifice reliability. Surprisingly, given the Propulsion Laboratory’s recommendation, the group postponed making any decision on the Rocketdyne effort until October.*63*

Reaction Motors made encouraging progress during September as the company continued to test the engine and injectors. The Rocketdyne program, however, failed to make any significant contributions, primarily because the company could not figure out how to mate its thrust chamber with the Reaction Motors ignition system. The X-15 Project Office conceded that the Rocketdyne effort was an "expensive and apparently fruitless" activity.*641

On 7 October 1958, the Technical Advisory Group reviewed the engine programs and concluded that although the Rocketdyne effort might offer higher performance at some point in the future, Reaction Motors was well on its way to producing an acceptable engine that would be available sooner. As a result, on 10 October 1958 the Propulsion Laboratory again recommended terminating the Rocketdyne effort, but this time Headquarters WADC and the X-15 Project Office agreed. Reaction Motors subsequently terminated the Rocketdyne subcontract.*651

Development progress continued at a reasonable pace during the remainder of 1958, despite several failures. For instance, Reaction Motors traced a destructive failure on 24 October to components that had already been recognized as inadequate. Since Reaction Motors was already redesigning the parts, the Air Force did not consider the failure significant.*66*

Despite the best efforts of all concerned during 1958, problems remained at the beginning of 1959. At a 20 January meeting of the Technical Advisory Group, Reaction Motors admitted the engine still suffered from injector failures at low power settings, excessive heat buildup during idle, and minor leakage from various components. A few days later, on 23 January, excessive vibration in a test engine at Lake Denmark resulted in a fuel-manifold failure. Despite the seemingly long list of deficiencies, it was apparent that the development effort would ultimately produce an acceptable engine.*67*

Static testing of prototype XLR99s and associated systems took place at the Reaction Motors facility in Lake Denmark, New Jersey. The test program used four test stands: three at Lake Denmark and stand E1 at the Picatinny Arsenal. The largest stand (R2 at Lake Denmark) was set up to test a complete aircraft system, including a structurally accurate aft fuselage, at all attitudes. Stands R2W and R3 at Lake Denmark were capable of horizontal firing only. The former was used for durability testing and environmental testing, and the latter was used for delivery acceptance tests because it was equipped with an elaborate thrust-vector mount. The test area at Lake Denmark contained support facilities with a storage capacity of 30,000 gallons of liquid oxygen, 18,000 gallons of anhydrous ammonia, and 4,000 gallons of hydrogen peroxide.

Reaction Motors began engine-system testing during the fall of 1958, and by the beginning of 1959 eight flight-representative engines were undergoing some level of testing. Engine run time progressed consistently, and the engines accumulated approximately 340 minutes of operation during the first quarter of 1959. Various components logged even greater run times, with the thrust chamber accumulating nearly 1,800 minutes and the turbopump over 4,200 minutes. The oxidizer pump, loosely based on the oxidizer pump used on the XLR30, operated at approximately 13,000 rpm. The fuel pump operated at 20,790 rpm and was essentially identical to the XLR30 unit. Each pump generated nearly 1,500 horsepower and had an output pressure of approximately 1,200 psi. The combined oxidizer/fuel flow rate at maximum thrust was 13,000 pounds per minute, exhausting the 18,000-pound propellant supply in 85 seconds.

The company finally reached a long-sought goal on 18 April 1959 when the first XLR99 completed its factory acceptance tests. This was the engine scheduled for use in the formal preliminary flight rating test (PFRT), which was based on an MIL-E-6626 modified to include "man-rating" requirements,. The completion of the PFRT series formed the basis of the engine’s approval for use in the X-15. The PFRT began the same day the factory acceptance tests were completed, and ran through 5 May 1960. The tests used four engines on test stands R2 and R3 at Lake Denmark, and E1 at Picatinny. Additional component tests took place at the Reaction Motors Component Laboratories and the Associated Testing Laboratories in Cadwell, New Jersey. Reaction Motors personnel conducted all of the tests under the watchful eye of Air Force engineers and inspectors. Captain K. E. Weiss, the XLR99 project engineer, was present for about half of the tests.[68]

The XLR99 was a great deal more complicated than the XLR11 engines used in most other X – planes. Reaction Motors conducted training classes for the Air Force and NASA personnel who would be responsible for operating and maintaining the engines at Edwards AFB. This was long before computer-aided instruction had even been dreamed of, and the classes were conducted using mimeographed course material and chalkboards. (U. S. Air Force)

In order to obtain a high level of confidence in the service life of the engine, the Air Force required two engines to each accumulate 60 minutes of operational time. Some of the tests were challenging: "[T]he engine shall be run at thrust levels of 50,000, 37,500, and 25,000 pounds for the corresponding durations of 87, 110, and 156 seconds. In addition, one run will be made at 90% of minimum thrust for 170 seconds duration and one run at 110% of maximum thrust for 80 seconds duration." In addition, to demonstrate the "all attitude" capability, an engine performed a series of tests while being fired with the thrust vector 90 degrees up and also 30 degrees down.1691

Unfortunately, the PFRT got off to a somewhat less than ideal start. The PFRT began with engine 012 performing the attitude test series. After it successfully completed nine 90-degree tests, Reaction Motors repositioned the engine for the 30-degree nose-down test. After several runs, a faulty weld in the second-stage igniter liquid-oxygen feed line developed a leak that resulted in a fire. The damage to the engine caused Reaction Motors to withdraw it from the test program for extensive repairs. To prevent further occurrence of this type of failure, engineers redesigned the igniter line to eliminate the weld, and the company revised its weld inspection program. The redesigned igniter line subsequently accumulated 1 hour of operation in engines 012 and 102 without incident. Since the original engine had not completed the 30-degree test series, all of those tests were repeated using engine 102.1701

Another problem was more serious, and continued throughout the flight program. During the PFRT, approximately 80 square inches of the Rokide Z171 ceramic coating used to insulate the firing chamber peeled off from engine 014. A heat-transfer analysis indicated that the loss of the Rokide coating would not produce a chamber burn-through, but the engineers did not understand why it came off. However, the engine successfully completed its 1 hour of operation, so Reaction Motors revised the acceptable Rokide loss specification based on this performance. Other problems included a transient vibration problem during start that could not be isolated. Fortunately, the built-in vibration cutoff circuit demonstrated that it would shut down the engine before a hazardous condition developed, and restarting the engine after the cutoff was usually successful. The test series experienced a variety of other minor problems, mostly resulting from faulty welds in various components, such as the turbine inlet and exhaust cases. The Air Force did not believe any of these were serious enough to terminate the tests or reject the engine.-1721

Reaction Motors conducted over 200 successful firings during the test program, accumulating 146 minutes of main chamber operation. In the end, one engine ran for 64 minutes and 100 starts; another ran for 65 minutes and 137 starts. The 231 seconds of specific impulse was 7 seconds below specification, but the engine met all other requirements. Engineers explained the low specific impulse by noting that "to expedite the development program, injector design was frozen before the optimum design was achieved." However, nobody expected the slight reduction in specific impulse to have any particular effect on the X-15 program.1771

Reaction Motors subsequently demonstrated the engine’s durability by accumulating more than 60 minutes of operating time on two different engines. One engine fired 108 times without having any more than routine maintenance. In addition, a series of 93 tests demonstrated that the engine would react safely under imposed malfunction conditions, and 234 engine tests demonstrated performance and safety requirements. Of these, 192 were full engine-firing demonstrations, and the remaining 42 were safety-limit tests that did not require thrust-chamber

operation. The PFRT cleared the engine to operate between 50% and 100% of full thrust. Testing continued, however, and the Air Force subsequently cleared the engine to operate at 30% of full thrust, meeting the initial contract specification.-1741

It is interesting to note that early in the proposal stage, North American determined that the aerodynamic drag of the X-15 was not as important a design factor as was normally the case with contemporary jet-powered fighters.-75 This was largely due to the amount of excess thrust expected to be available from the engine. Engineers considered weight the largest driver in the overall airplane design. Only about 10% of the total engine thrust was necessary to overcome drag, and another 20% was required to overcome weight. The remaining 70% of engine thrust was available to accelerate the X-15.76

At the time it was built, the XLR99 was the largest man-rated rocket engine yet developed. Of course, this would soon change as the manned space program accelerated into high gear. The 915-pound XLR99 could produce 50,000 pounds of thrust (lbf) at sea level, 57,000 lbf at 45,000 feet, and 57,850 lbf at 100,000 feet. The nominal oxidizer-to-fuel ratio was 1.25:1, and the engine had a normal chamber pressure of 600 psi. Playing with the oxidizer-to-fuel ratio could slightly increase the thrust, and the amount of thrust varied somewhat among engines because of manufacturing tolerances. Some engines produced over 61,000 lbf at specific altitudes. The engine had a specific impulse of 230-lbf-sec/lbm at sea level and 276-lbf-sec/lbm at 100,000 feet. The engine was throttleable from 30% to 100%, although the first couple of engines were limited to 50% on the low end until early 1962. Even after Reaction Motors modified the engines and the Air Force approved the use of 30% thrust, a high vibration level meant that they were operationally restricted to no less than 40% thrust. The amount of available propellant was all that limited the duration of any given run. Reaction Motors estimated the service life (mean time between overhaul) of the engine at 1 hour or 100 starts.77

At the beginning of the X-15 program, researchers used the methods developed by Edward Van Driest and Ernst Eckert to determine the heat-transfer coefficients for temperature calculations. However, the measured heat-transfer coefficients during the early flight program were considerably lower than the predicted values. Based on these preliminary results, derived primarily from the initial low-angle-of-attack flights, engineers modified Eckert’s turbulent-flow method to produce the adiabatic-wall reference-temperature method.[48]

fuselage. The boundary-layer transition was completely unpredictable, but since researchers expected turbulent flow during the major portion of most flights, they normally used turbulent – flow calculations for the entire flight. Next came determining the heat-transfer coefficients, and finally calculating the skin temperature. Due to the tedious work involved in this process, which was done mostly by hand since general-purpose computers were not yet in widespread use, the researchers made many assumptions that simplified the procedure. For instance, it was assumed that temperature did not vary through the thickness of the skin, no heat was transferred along the skin, the specific heat of the skin was constant, solar radiation to the skin was negligible, the emissivity of the skin was constant, and no net heat transfer occurred between surfaces by radiation.[49]

Temperatures calculated using the adiabatic-wall reference-temperature method tended to agree closely with measured data from the flight program. In several instances the calculated temperature was somewhat higher because the analytical method assumed turbulent flow all of the time. This was considered reasonable and sufficient for flight-safety purposes since it erred on the side of caution.-50

In 1957, Lockheed Aircraft Company developed a thermal analyzer program that ran on an IBM 704 digital computer, the largest of its type then available. This program was capable of running the heating prediction equations, including the effects of transient conduction, convection, radiation, and heat storage, that researchers had previously omitted for the sake of expediency. With Lockheed’s assistance, researchers modified the program to reflect the X-15 configuration. The program estimated the heat input to the skin elements using the attached-shock Prandtl- Meyer expansion method for flow conditions, and the adiabatic-wall reference-temperature method for heat transfer. Researchers used the laminar-flow theory of Fay and Riddell to compute the heat input to the stagnation points, with curves developed by Lester Lees used to weight the periphery.-50

One of the primary goals of the X-15 program was to validate the various heat-transfer methods with actual flight results. Many of the early X-15 flights were dedicated to gathering data that the researchers would spend years comparing against wind tunnel and theoretical results. The results were vastly improved heat-transfer models that were used during the Apollo and space shuttle programs. (NASA)

To accompany the Lockheed-developed software, North American developed two other programs to predict structural heating values and their distribution along the airframe. The first program computed local-flow conditions on the aircraft, and the second program used the local-flow conditions to calculate the aerodynamic heat transfer to the skin. The program developed by Lockheed calculated the transient heating of internal structure based on the results of the other two programs.-1521

To evaluate the acceptability of the thermal analyzer program, researchers compared calculated results with actual flight results on several occasions. The values always compared favorably, and were usually slightly better than the hand-calculated values for the same conditions. North American and NASA quickly adopted the automated process based largely on the tremendous labor savings it offered.

After the flight planners established a flight profile on the fixed-base simulator, they digitized the results of a clean flight and input them into the IBM 704 to predict the skin and structural temperatures and thermal gradients for the flight. This was a time-consuming process. Researchers then compared the resulting data with the design conditions to ensure that the X-15 did not violate any structural margins. If any exceptions were uncovered during the comparison, researchers modified the flight profile and the entire process was repeated. Emergency and contingency flight profiles went through the same rigorous process. After the flight, researchers compared the heating predictions with actual flight data and then refined the simulations.-1531