First-Hand:The X-15 Project - Design, Construction, and Preparation - Chapter 11 of the Experimental Research Airplanes and the Sound Barrier: Difference between revisions
| Line 147: | Line 147: | ||
Initially, the Navy had rigged the centrifuge to simulate flight g-force profiles in a programmed manner with precut cams, but what the X-15 project needed was a way of varying the acceleration forces to represent airplane dynamic responses to pilot control actions. What this took was interposing an analog computer simulation of X-15 flight dynamics, that took pilot inputs from a replica X-15 cockpit in the gondola, and in turn computed control orders to vary centrifuge speed and gondola orientation. The mechanization was considerably more difficult than the Iron Bird simulator, but it worked. X-15 pilots were dispatched to Johnsville and made about 400 simulated flights long before the first real X-15 flight. Many of these flights simulated reentry's with failed systems, or unusually high angles of attack or yaw angles. In most cases the accelerations were more severe than actually experienced in later X-15 flights, and the testers found that the side located wrist controller worked well under the worst of conditions, in fact up to 12 gs. The centrifuge flights also made useful contributions to improvements of the side controller, the pilot’s restraint system, and instrument displays.[56, Chapter 4] | Initially, the Navy had rigged the centrifuge to simulate flight g-force profiles in a programmed manner with precut cams, but what the X-15 project needed was a way of varying the acceleration forces to represent airplane dynamic responses to pilot control actions. What this took was interposing an analog computer simulation of X-15 flight dynamics, that took pilot inputs from a replica X-15 cockpit in the gondola, and in turn computed control orders to vary centrifuge speed and gondola orientation. The mechanization was considerably more difficult than the Iron Bird simulator, but it worked. X-15 pilots were dispatched to Johnsville and made about 400 simulated flights long before the first real X-15 flight. Many of these flights simulated reentry's with failed systems, or unusually high angles of attack or yaw angles. In most cases the accelerations were more severe than actually experienced in later X-15 flights, and the testers found that the side located wrist controller worked well under the worst of conditions, in fact up to 12 gs. The centrifuge flights also made useful contributions to improvements of the side controller, the pilot’s restraint system, and instrument displays.[56, Chapter 4] | ||
Click here to proceed to Chapter 12 of the Experimental Research Airplanes and the Sound Barrier - The X-15 Project - Flight Testing | [[First-Hand:The X-15 Project - Flight Testing - Chapter 12 of the Experimental Research Airplanes and the Sound Barrier|Click here to proceed to Chapter 12 of the Experimental Research Airplanes and the Sound Barrier - The X-15 Project - Flight Testing.]] | ||
Revision as of 19:09, 9 August 2016
By David L. Boslaugh, CAPT USN, Retired
Inception of the X-15 Project
The conception of the Mach 3 capable X-2 began as early as 1945, spurred on by visits of American aeronautical experts to Germany where they were astounded by the advances the former enemy had made in high speed flight. In particular, they noted the German discoveries of the low drag features of swept wings, that were later incorporated in the X-2. As early as 1951, visionaries were already thinking of the need for a research airplane that could fly at least another Mach number faster than the X-2. They could even see a need for aerodynamic studies beyond Mach 5 (defined as the lower boundary of hypersonic flight) to support high speed missile development; in particular intercontinental ballistic missiles. [26, p.2, 7] NACA and other government agencies were already running wind tunnels at speeds beyond Mach 5. In fact the Naval Ordnance Laboratory at White Oak , Maryland, had removed the uncompleted Mach 10 tunnel from the German’s Peenemude Laboratory, and had it in operation by 1950. The Germans had actually tested a winged version of their V-2 rocket up to a speed of Mach 4 when it disintegrated. [26, pp.9-10] What was needed was a research airplane capable of the same speeds to validate wind tunnel findings, and adjust wind tunnel design and processes if necessary.
During 1951 and early 1952 Hubert M. Drake, Assistant Chief of the HSFS Research Division, and engineer L. Robert Carman worked on a feasibility study and a proposal to modify the forthcoming X-2 research airplane with more powerful engines to achieve speeds up to Mach 4.5. The HSFS submitted the Drake/Carman proposal airplane to NACA headquarters in 1952. [17, p.226] Others proposed modifying a Douglas D-558-2 Skyrocket with thinner wings to achieve higher speeds. In the mean time, Robert J. Woods, chief design engineer of Bell Aircraft Corp. who had designed the X-1 series, proposed at the January 1952 meeting of the NACA Committee on Aerodynamics that NACA should build a manned hypersonic airplane capable of flights out of the atmosphere. He noted that there were many unknowns regarding aircraft structures, heating, stability, and control in achieving space flight that could be resolved with such a craft. NACA officials replied they would take the three proposals under advisement. [26, p.11]
In February 1954 the NACA Interlaboratory Research Airplane Projects Panel, meeting in Washington, D.C., was directed to review and comment on the proposals for a hypersonic research plane. The Panel first went back in history to review the contributions of preceding high speed experimental research airplanes. They noted that although the X-1 had been the first to exceed the speed of sound, and proved it could be done, the X-1 series had barely stayed ahead of the rapidly advancing technology being incorporated in operational fighters. The XS-1 breached the sound barrier in October 1947, but there is even a claim that the prototype North American F-86 Sabre fighter had exceeded the speed of sound in a dive two weeks before. By 1949, Air Force pilots were routinely diving the F-86 into the supersonic region. [74] Further, the Republic XF-91 experimental fighter, the first U.S. fighter specifically designed to exceed the speed of sound, first flew in May 1949, and by December 1951 it had gone supersonic in level flight. [78] Scott Crossfield had been the first to fly twice the speed of sound on On 20 November 1953, but by February 1956 the Mach 2 capable Lockheed YF-104 Starfighter was in the air. [71] Considering the above, and the resultant criticism of preceding research airplane programs, the Panel agreed that the proposed research plane should provide a substantial advance in performance beyond present and planned research airplane capabilities. [26, p.14]
The Panel decided that modifications to the X-2 or to other existing research airplanes were not feasible due to their inability to survive expected aerodynamic heating at speeds above Mach 3, and instead sent forward a recommendation to NACA Headquarters that a new hypersonic research airplane project capable of speeds up to Mach 10 and flights above the atmosphere should be started. The panel also proposed that each NACA lab should prepare recommendations for the capabilities and research goals for such a plane. [26, pp.11-12] By March 1954, each of the labs had set up small informal hypersonic research study groups, and the Langley Laboratory in particular put considerable effort into the study including a proposed design. One of the problems confronting a hypersonic airplane that had the capability to rise above the atmosphere would be severe aerodynamic heating on reentry, and they had used their hypersonic wind tunnel to devise a possible solution. Their solution was to reenter the less dense higher atmosphere at a high angle of attack to begin energy dissipation early at relatively low levels of heating. Further, they had been warned by HSFS that the X-1s at high angles of attack at supersonic speeds would significantly lose directional stability. Their solution to that was a fairly thick, diamond shaped all moving vertical stabilizer. Langley had confirmed in hypersonic tunnel testing that a stabilizer formed by ten degree wedges could be made small enough to be effective in high angle/high drag reentrys. They also proposed a cross shaped tail assembly with a ventral vertical stabilizer extending as far below the fuselage as the dorsal stabilizer did above the fuselage. This was to keep the airplane’s inertia axis in line with the flight axis to help prevent the problem of roll coupling. They proposed that the craft should be powered by three or four rocket engines, and that it be air launched as were the preceding rocket research airplanes. [17, pp;.227-229]
Langley also grappled with the problem of how to cope with high structural heating on reentry. They rejected the idea of a high temperature outer shell backed by some form of insulation in favor of a heated internal structure made of a heat resistant metal. They chose the steel-nickel-chromium alloy Inconel X that could retain its strength at temperatures over 1,200 degrees F to be used for the skin and critical structural members. They had found, fortuitously, that the skin thickness needed for the required structural strength was just about the right mass to keep the craft from overheating on reentry. They also came up with segmented structural design features to cope with expected unequal thermal structural expansions due to temperature differentials. [26, pp.2-13]
In July 1954 NACA Administrator Hugh Dryden convened a meeting with Air Force and Navy aeronautics research officials where he outlined a proposed NACA hypersonic research airplane project. Both services revealed that they had already been considering such a project, and agreed to a joint NACA/Air Force/Navy program. Also, as with previous rocket powered research airplanes, they agreed that the craft should not be burdened with any requirements for a future operational military airplane. [17, p.229] The Langley study had noted that there was no one rocket engine capable of powering the proposed research airplane, and had proposed use of four clustered General Electric Hermes rocket engines. However, the Air Force Wright Field Power Plant Laboratory noted that the Hermes engine was not “man rated”, and recommended that engine selection be deferred pending further study of engine requirements. The principal Navy Bureau of Aeronautics review recommendation was a fairly curious one, that the aircraft should have accommodation for a human observer instead of the normal NACA instrumentation package. In a 22 October meeting of NACA and the Air Force engineers it was agreed that development of a suitable rocket engine for the research plane would be made a separate project from the research airplane, and that the engine would be government furnished to the airframe contractor. They also agreed that three of the proposed airplanes should be acquired. [26, pp.13-14, p .19]
In late December, senior NACA, Navy and Air Force officials met at NACA headquarters to draft a memorandum of understanding. It was agreed that: NACA would have technical control over the project, that the two services would fund the design & construction stages, that NACA would be in charge of planning the test program, flight testing, and reporting results. It was also reconfirmed that the airplane would be purely a research aircraft. The Navy and Air Force also agreed to each provide two engineering test pilots to the project. Previous service sponsored experimental research airplanes had first been delivered to the sponsoring service for initial flight testing before being turned over to NACA, however the X-15s would be turned directly over to NACA after contractor demonstration flights. [7, p.286] It was further agreed that NACA would maintain and operate the research airplanes, except that the Air Force would maintain the engines, the pilot’s pressure suits, and the ejection seat. NACA would also maintain and operate a new flight test tracking range, a flight simulator, and half of the chase planes. [58, p.83]
At the end of December, the Air force sent out specifications and invitations to bid (IFB) to 12 contractors. The specification called for a manned research airplane capable of surviving stresses and heat load of speeds up to Mach 6 and altitudes up to 250,000 feet (Approximately 47 miles). It further called for accommodating 800 pounds of NACA instrumentation in a volume of 40 cubic feet with the instrumentation consuming 2.25 kilowatts of electric power. As per the Navy requirement, an engineering study of an alternate two-place design with an observer in place of the instrumentation was called for. With respect to the rocket engine, the bidders were free to propose an engine, however if it was chosen, the Air Force would acquire it and make any necessary modifications. Langley Laboratory’s study paper was included in the IFB, but it was for information only and not a part of the specifications. The three planes were to be delivered within 30 months after date of contract. [26, p.20]
Four bidders responded: Bell Aircraft, Douglas Aircraft, Republic Aviation, and North American Aviation (NAA). During May, June, and July 1955, NACA, Air Force, and, Navy source selection teams independently evaluated the proposals, and concluded that the North American proposal was the best even though the NAA cost proposal considerably exceeded the Government cost estimate, and eight months would have to be added to the 30-month schedule. On 11 June 1956 the Air Force awarded North American Aviation a contract for three hypersonic research airplanes to be given the Air Force designations: X-15-1, -2, and -3. North American proposed a B-36 intercontinental bomber with six reciprocating and four jet engines as the mother airplane because they calculated that a B-50 bomber could not lift the X-15 to the required 30,000-foot launch altitude. [26, p.20, 28, 32] The contract called for NAA to do initial flight testing to demonstrate general airworthiness and specifications satisfaction. Their testing, called Phase I, was to cover: stability and control at positive and negative gs, engine reliability, satisfactory in -flight launching ballistic controls, structural integrity at high gs, and safe landings. The contractor was not required so demonstrate performance in excess of Mach 2 or 100,000 feet altitude. [7, p.285] The first X-15 was to be delivered on 31 October 1958. [26, p.33]
Scotty Crossfield Leaves NACA
By late 1955 it was clear that the X-15 contract was going to be awarded to North American Aviation, pending a few formalities. This prompted senior HSFS test pilot Scotty Crossfield to make a proposal to Station chief Walter Williams that he should be assigned as NACA liaison engineer at the NAA plant. Williams demurred on the grounds that such an assignment could be construed as NACA interference with their work. He also reminded Scotty that he would eventually be flying the X-15s when they were delivered. Crossfield wanted so badly to be a part of the X-15 design and construction process that he was willing to give that up, so he tendered his resignation to NACA and made job application to North American Aviation. In December 1955 he was hired as prospective contract demonstration pilot, and in the mean time he would serve as general project consultant. As time went on, Crossfield and North American’s chief engineer Charles Feltz decided on Scotty’s principal function. As a research project never before done, there were going to be a myriad of forces and people trying to make airframe and system design changes, and his function was to challenge and verify each one, with the goal of keeping the design as simple, stable, and workable as possible. He was to be the X-15 project’s “chief son-of-a-bitch.” [7, pp.209-212, p.225]
Design and Construction of the X-15
The Airframe
North American concurred with the Langley Lab study paper’s “hot structure” design with no external insulation (other than around the cockpit). Initially the NAA design called for heat resistant ablative material that would be worn away to protect hot spots, such as wing leading edges, but this was abandoned in favor of more massive amounts of a heat resistant metal at the hot spots. [7, p.229] In a search for a strong, highly heat resistant metal, NAA settled on Inconel X that had been proposed in the Langley study. The metal was extremely tough and would hold its strength up to air friction temperatures well above 1,200 degrees F. All of the airplane’s skin would be Inconel X, and most of the internal wing and fuselage load carrying structures would be of titanium except for high heat areas that would be Inconel X. Internal structures not subject to high heats and high loads would be aluminum. [58, p.44] With respect to withstanding the structural stress of high g maneuvers, the X-1 and D-558-II series had been stressed to withstand 18gs, and the X-2 stressed to 12 gs, however they had never been subject to much more than 7 gs. The X-15’s longitudinal normal acceleration design limits took advantage of this experience, and were relaxed to plus 7.33 gs and minus 3gs at 30% propellant weight, and +4 and -2 gs at full propellant weight. [52, p.83]
Preceding rocket research airplanes had used liquid oxygen and alcohol for propellants, however X-15 project mangers decided to use anhydrous ammonia in place of alcohol because of its greater energy content. Anhydrous means ‘without water’, or in other words pure undiluted ammonia, a much more dangerous substance than alcohol. Liquid oxygen tank capacity would be 1,038 gallons, and the ammonia tank 1,445 gallons. As in the later X-1 series, hydrogen peroxide, decomposed to high pressure steam, would drive a turbopump feeding propellants to the rocket engine. The main peroxide tank would hold 77.5 gallons, and there would also be smaller hydrogen peroxide tanks. These would provide steam to auxiliary power units turning electrical generators and hydraulic pumps, and would also supply peroxide to the small out-of-atmosphere reaction control system steam jets. [26, p.183] Helium would be used to pressurize the oxygen and ammonia tanks to drive the propellants into the engine turbopump, and would also be used to expel propellants if they had to be jettisoned in flight. As with earlier research airplanes, the X-15 would be too heavy to land with propellants aboard, and the specification required that a full load of propellants had to be jettisoned within two minutes in case of engine malfunction. The main high pressure helium tank would form the core of the liquid oxygen tank, and there would be smaller helium tanks located where space allowed. Previous rocket research planes had been built with propellant tanks inside, and separate from, the airplane’s outer skin, but the X-15 propellant tanks outer walls would be formed by the Inconel X skin, allowing a considerable weight saving. [58, p.35, 46]
The cockpit would be a sealed aluminum structure within the Inconel X skin, and it would have a layer of insulation protecting the pilot from the expected intense skin heat. The pilot would wear a full pressure suit in case of cockpit depressurization, but the cabin would normally be pressurized to 35,000 feet to keep pilot’s pressure suit from inflating except in emergencies. Pure nitrogen would be used to pressurize the cockpit to insure there would be no oxygen to sustain an in-flight fire. There was a sober consideration in using a pure nitrogen environment. Even though the air we breathe is 78% nitrogen, that gas alone will not sustain life, and what’s worse, just a couple breaths of pure nitrogen will cause cause rapid asphyxiation. The pilot therefore would need his own constant supply of breathing oxygen. [58, pp.37-38] Scotty Crossfield writes of an occasion when running ground tests in the X-15 cockpit with pressure suit and helmet on. He accidentally lost his oxygen supply and began breathing pure nitrogen coming from a crack in his pressure suit neck seal. In seconds he began to succumb, and it was only thanks to an alert ground crew man who saw his distress and opened his face plate that he survived. [7, pp.329-330] There would therefore be compressed nitrogen tanks to provide cockpit pressurization, cooling of the pilot’s pressure suit, cooling of the auxiliary power units’ bearings, and cooling for a special yaw and angle of attack sensor in the nose. [58, p.43]
The rocket propelled Douglas D-558-II was designed to confine all piping, wiring and control cables within its circular fuselage, but the Bell X-1 and X-2 designs featured “tunnels” along the fuselage top and bottom to enclose these items. North American engineers planned to use tunnels similar to the Bell design, however, during the evolution to the final shape of the X-15 airframe, wind tunnel testing found that “maintenance tunnels” along the top and bottom of the fuselage would destabilize the plane in yaw. North American engineers ran the tunnels along each side of the fuselage instead, and found in wind tunnel tests that the side tunnels not only cured the yaw instability but also increased lift & improved propulsion efficiency. [7, p.229] It was found that at high Mach numbers the side tunnels would provide almost half the aerodynamic lift. Overall, wind tunnel testing on X-15 models consumed more than 4,000 hours in 15 different wind tunnels. [26, p.35, 37] More wind tunnel tests showed that instead of the wide diamond shaped vertical stabilizers proposed in the Langley study, a shape continuing the ten degree front-half wedge all the way to a wide, blunt trailing edge would be even more effective. [17, p.235]
The X-15 wing span would be only 22 feet and wing area would be 200 square feet. [58, p.40] For comparison, the Piper Cub’s corresponding dimensions are 35 feet, and 179 square feet. Wing cross section would be a modified thin NACA 66005 airfoil with 5% thickness. It would have titanium internal structure and Inconel X skin. The wing and vertical stabilizer leading edges would be made of heavy machined Inconel X bar stock, and the leading edge of horizontal stabilizers formed from Inconel X sheets would be built up internally with weld material to form a heat sink. As recommended in the Langley studying, wing leading edges would be divided into segments with expansion joints to accommodate thermal expansion. [26, p.181] Wing skins would be tapered from 3/16-inch at root to 30 thousandths of an inch at the tip, and the skins would be milled from Inconel X blanks by grinding, because the alloy was so tough it blunted normal milling tools. Welding, shaping, and fabricating Inconel X required much experimentation; it was a research project in itself. Before the X-15s, the alloy had only been worked in a laboratory environment, and there were no applicable handbooks or industry guides. For example, it was found each weld had to be heat treated in ovens to relieve welding stress, and NAA had to make seven different wing skins for the first X-5 before achieving a satisfactory welding process. In this work, the company was a pioneer who provided much valuable information to industry. [7, pp.291-292] A black body radiates heat better than any other color, so the X-15s would be painted black on delivery. However, as the skin experienced high heating flights and weathered, the Inconel X itself would produce a smooth blue/black oxide layer that was even better than the paint. [15, p.16]
The X-15 specification was not explicit concerning pilot emergency escape provisions, however, Air Force policy at the time called for all new USAF airplanes to be equipped with escape capsule such as had been built into the D-558-II Skyrocket and X-2 research airplanes. Scotty Crossfield had flown the Skyrocket numerous times, and after studying its escape capsule design and function, he was convinced that if the capsule was separated at supersonic speeds, it would decelerate so rapidly when free in the air stream that the pilot would be incapacitated. Crossfield swore he would never use the capsule, but would remain with the plane until it decelerated to safe airspeed and he would bail out in conventional manner. He also swore he would build a case to obviate the use of an escape capsule in the X-15, in favor of an ejection seat. In his resultant study he was able to prove that a capsule would increase X-15 weight by 9,000 pounds, would stretch the schedule by more than a year, and would be more dangerous than an ejection seat. He was able to get the Air Force to reverse its policy for the X-15. [7, pp.230-233]. The resultant ejection seat was designed with stabilizing vanes that would extend to keep the seat aligned with the relative wind, and would be useable at speeds up to Mach 4 and up to 120,000 feet. [58, p.38]
A ride in an experimental research airplane often threw the pilot rather violently around in the cockpit, sometimes to the point of being knocked unconscious, even with tight seat straps. In their study of what would make a comfortable seat that could hold a pilot in place, Crossfield and the North American engineers asked themselves, “What vehicle has a seat designed for comfort in rough riding conditions?” Their answer was farm tractors, and they consulted International Harvester Company engineers who had done considerable study of tractor seat comfort, including determining the human spine’s natural frequency. The X-15’s pilot seat bottom thus turned out to have the same shape and resiliency as a tractor seat, with considerable saving in weight. [7, pp.244-245]
The X-15’s wings would be too thin to house retractable landing gear, and there was no room in the fuselage for landing gear. The Bell X-2’s main gear was a single center-line skid that had caused some landing instabilities, so the X-15s would be fitted with two skids under the tail. They would be retracted flush with the fuselage bottom in flight. Just before landing the pilot would release them with a pull cable, and a combination of airflow and gravity would bring them down to a locked position. They would have no retracting mechanism, and would have to be manually retracted on the ground. The nose gear would be a double wheel non-steerable arrangement retracted into a well in front of the cockpit, and it would also be lowered, when released, by gravity and air flow. As with the rear skids, the nose gear would be manually retracted on the ground. There would be no brakes, and the main slowing force on landing would be friction of the rear skids with the lake bed, usually needing more than a mile to stop. [39, p.126]
The Auxiliary Power Units
The X-1 series and the Douglas Skyrockets used pilot actuated control systems with cables, push rods, and bell cranks to position the control surfaces, and they used batteries to power the electronics and instrumentation. The X-2 had a fairly simple hydraulic system to power-boost ailerons and elevator, but used conventional cables and bell cranks from the pilot’s rudder pedals for rudder control. This writer has not been able to determine the power source for the X-2’s hydraulic system, but it might have been as simple as a nitrogen-charged hydraulic accumulator. Expected control forces at hypersonic speeds were going to be far beyond muscle power, and the X-15s were going to need hydraulic control systems calling for powered hydraulic pumps. Since some sort of mechanical power unit was needed, it was appropriate that the same unit should turn a generator for electrical power. North American decided on a turbine power supply driven by high pressure steam generated by decomposed hydrogen peroxide passed over a silver mesh catalyst. The term for them would be auxiliary power units (APUs) even though they would be far from “auxiliary”; they would be the primary source of electrical power, and the only source of hydraulic power. The planes would also have a 24-volt emergency battery. Each plane would be fitted with two APUs, each turning both an eight kilowatt generator and a hydraulic pump. [7, p.311] [26, p.187]
Each of the APUs would provide half of the plane’s maximum needed power, and if one APU failed, the remaining would provide enough power to complete an emergency landing. If both APUs failed, the craft would have no hydraulic power and no instruments; the pilot would have no choice but to bail out. [7, p.313] Each APU had its own hydrogen peroxide tank with 45 minutes of fuel. The mother plane’s climb to launch point could take up to two hours so, when captive, the X-15 could take electrical power either from the B-52 or from the APUs. Normal practice would be to start the APUs about 12 minutes before launch. [58, p.39] Auxiliary power units such as needed by the X-15 had never before been built so there was nothing suitable off the shelf. Each unit had to produce over 40 horsepower, but had to be small and light. Design was going to be on the fringe of what was physically possible, and North American turned to a company that specialized in power units of all sorts, General Electric Company. [7, p. 313] The APUs would be a maintenance and reliability challenge throughout the X-15 project life, but with constant GE tweaking and redesign, they would gradually improve. [26, p.97] As an example problem, they would have a tendency to fail more under the extremely low atmospheric pressures of high altitude flight due to lubricating oil forced out of the bearings. The High Speed Flight Station solution was to enclose them within a box pressurized by nitrogen. [15, p.58]
The XLR99 Rocket Engine
North American had proposed using a Reaction Motors XLR30 rocket engine that was well along in development, and this was approved by the selection committee, with the proviso that the Air Force would acquire the engines under a separate contract. In September 1956, the Air Force awarded Reaction Motors a contract to deliver one XLR30 rocket engines for Air Force testing. After testing, the Air Force fed back required changes to be incorporated in ten production engines to be designated XLR99. [26, p.33, 49] The XLR99 would be able to provide 57,000 pounds of thrust at 40,000 feet, whereas the fully loaded X-15 weighed 33,000 pounds. [58, p.35] This means the plane could stand on its tail and fly straight up. This would not be a good idea , however, because if the craft left the atmosphere going straight up with no horizontal velocity component, it would then fall straight down until reentry. If reentering going straight down, it was doubtful the pilot could pull out before ground impact.
Air Force specifications required that the engine be throttleable from 30% to full thrust, and that it be stoppable and restartable a number of times. The propellant tanks would be pressurized at 50 psi by helium to push the propellants into the engine turbopump, and throttle control would be by pilot varying the speed of the turbopump. The propellants would circulate around the outside of the combustion chamber to cool it before being forced into the combustion chamber operating at a pressure of 600 psi. [58, p.45] At full thrust, the propellant load would last 88 seconds. [7, p.223] The engine would have a two-stage ignition system beginning with a small mall igniter, similar to blow torch fed by a small stream of propellants and lit by a spark. The small first stage igniter would be lit before drop and would run continuously. This was called running the engine in idle. To start the engine, propellant would be forced into a second stage igniter, which would be ignited by the flame from the first stage igniter. The second stage igniter was, in itself, a small rocket engine producing 1,500 pounds of thrust. The second stage igniter would then ignite the main propellant flow. Main propellant flow could be up to 30 gallons per second. [58, p.45, 78] The engine with turbopump weighed 1025 pounds, overall length 82 inches, and nozzle diameter 39 inches. [39, p.128]
Control System Design
As already mentioned, the outboard sections of the X-15s wedge shaped vertical stabilizers mounted above and below the fuselage would be all-moving. The thin, five percent thick, horizontal stabilizers would also be all-moving. The horizontal tail surfaces would serve not only for pitch control, but would be moved differentially for roll control, that is the leading edge of one stabilizer would go up and the other down to effect roll torque. This was called a “rolling tail.” The X-1 and D-558 series research airplanes had direct mechanical linkages between the pilots controls and the control surfaces, however the X-2s used hydraulic actuators to position the elevator and ailerons, though the rudder was positioned by cables connected to the pilot’s rudder pedals. As with the X-2 elevator and ailerons, forces required to position the X-15’s aerodynamic control surfaces at hypersonic speeds were going to be far greater than a human could muster. The plane would use hydraulic actuators to position all control surfaces and the wing flaps. The pilot would not get any feedback force from the deflected control surfaces; instead he would get control “feel” by springs holding the control stick and rudder pedals at central position. The pilot’s controls would be linked to the hydraulic actuators by a system of cables, push rods and bell cranks. An arrangement of levers and bell cranks combined pilot’s roll and pitch control inputs to move the horizontal stabilizers in unison for pitch control and the same time differentially for roll control. [60, p.3-4] The plane would have two separate hydraulic systems, each with its own hydraulic cylinder actuators, and each providing half the force to move control surfaces. One system by itself would be powerful enough to control the plane in an emergency. The hydraulic system also sent pressure to position the airflow direction sensing ball in the nose as well as the speed brakes. The brakes would form the rear half of the inboard sections of the upper and lower vertical stabilizers. [26, p.187]
The X-15s were designed to ascend to over 300,000 feet, whereas aerodynamic control surfaces were only effective at altitudes below about 150,000 feet. The means of controlling aircraft attitude in the region of aerodynamic control ineffectiveness would be by small hydrogen peroxide rockets located in the wing tips and nose. The concept had already been evaluated by test pilot Neil Armstrong in four flights of the X-1B fitted with the prototype reaction control system. As with the hydrogen peroxide driven main propellant pump and the auxiliary power units, pressurized hydrogen peroxide would be passed over a silver mesh catalyst, causing the peroxide to decompose into a jet of high pressure steam, creating up to 100 pounds of thrust. For “fail safe” purposes, the reaction control system would be duplicated with two up and two down rockets at each wing tip and in the nose and also two right and two left in the nose. Pilot control would be by a handle mounted on the left side of the cockpit. Pushing the handle up or down would raise or lower the nose, pushing it left would move the nose to the left, and rotating the handle would roll the plane in the direction of handle rotation. Using the early versions of the system would take considerable pilot skill to position the craft where the pilot wanted it in the frictionless conditions of near space. [58, p. 37] [26, p.187, 190]
The preceding research airplanes, as well as wind tunnel testing of X-15 shapes, showed that there would be flight regions where static pitch and yaw stability (the tendency to weathervane into the air stream) and/or damping of motions about the three rotational axes (the “fly swatter” resistance to motion effect) would be almost nonexistent. This could lead to dangerous flight divergences such as the X-2 disaster at Mach 3, and to oscillations beyond the ability of pilots to control. In some instances, attempts at pilot control could even make them worse by pilot induced oscillations. It was essential to provide the X-15s with a stability augmentation system (SAS) that would enhance natural aerodynamic damping about all three axes at some flight conditions. [7, pp.338-339] Such a system would have rate gyros that sensed rotation about the three axes, and would send signals to apply control surface deflections to oppose the rate and direction of rotation. The signals would actuate small hydraulic cylinders that added or subtracted control inputs to a bell crank linkage that combined pilot inputs and SAS inputs to the main hydraulic control servo cylinders. [60, p.3-4] The stability augmentation system would have two redundant control channels and a difference detector that would shut down a particular axis if the difference in the two channels exceeded a given threshold. In the first two X-15s, the pilot would have a ten-position selector switch to set damper gain to suit a particular flight condition. After experience, it was found, as a simplification, that a gain selector having just a high and a low position would be just as effective because they found they were usually calling for a lot of gain or relatively low gain. [26, p.189] The SAS in the first two planes would be provided by Westinghouse Electric Company. An advanced, much more sophisticated system, would be provided for X-15-3 by Minneapolis Honeywell Regulator Company, and will be discussed in more detail later.
Flight Test Instrumentation
As requested by NACA, the Air Force specification called for weight and space reservation for 800 pounds of flight test instrumentation to be installed by NACA after delivery. However, in mid 1957 NACA asked the Air Force to modify the contract to increase instrumentation load from 800 to 1,500 pounds. This was an unpleasant surprise to North American engineers who had to pare structure weight in many places to accommodate the additional instrumentation weight. [7, pp.262-263] The literature is silent on what prompted NACA to ask for the additional instrumentation. As delivered, the X-15s accommodated the following instrumentation:
- 656 thermocouples and 112 strain gages molded and baked permanently in the skin at various locations - done by NAA before delivery
- 140 pressure sensors drilled in the skin at various locations with tubing run to the instrument bay - done by NAA before delivery
- 90 telemetering system pickup sensors
- 22 flight parameter sensors such as yaw angle, angle of attack, air speed, and altitude
- 28 control surface position indicators
- 212 airplane condition monitors
- 8 pilot biomedical measurements
- 36 channel oscillograph film recorder plus other film recorders
- camera behind pilot’s right shoulder to record the instrument panel
[26, p.158]
Sensors such as thermocouples, strain gages, pressure holes, and control position indicators, were located throughout the craft, and transmitted their measurements to recorders in a pressurized instrument bay behind the cockpit. The devices in the instrument bay were mounted in a rack that could be detached and lifted out of the plane for easy access. The instrument bay was large enough that it could have accommodated a second crew seat as the Navy had requested, however, this was never done. There were also smaller instrumentation compartments in the nose, and at the plane’s center of gravity. The latter was especially important for housing linear accelerometers so that they would not sense extraneous accelerations due to airplane rotations. To measure angle of attack, yaw angle, static air pressure, and ram air dynamic pressure, previous research airplanes had been fitted with an instrumentation boom extending out in front of the plane’s nose. It was called a YAPS head, standing for Yaw angle, angle of Attack, Pitot tube, and Static pressure. Yaw angle and angle of attack were measured by small vanes in the air stream. The problem was, at speeds above about Mach 3.5, these small vanes would be melted by aerodynamic heating, so a different approach would be needed for hypersonic flight.
The solution was called either the ball nose or the Q-ball. Q being the aeronautical engineering symbol for dynamic ram air pressure. The Q-ball was a hollow 6.5-inch Inconel X sphere mounted at the front of a 13-inch long Inconel X cone. The sphere was articulated and could be positioned in pitch and yaw attitude by a small hydraulic servomechanism. Four pressure sensors were mounted in a square pattern on the front of the ball with a dynamic pressure sensor at the center of the pattern. The hydraulic servomechanism measured the pressure differentials between the right and left sensors, and the high and low sensors, and positioned the ball so that the pressure differentials were zero. The result was the dynamic pressure sensor would be facing directly into the air stream, and the ball’s right/left and up/down angles would be measures of angle of attack and yaw. Static ambient air pressure would be picked up at a sensor in the fuselage skin. The Q-ball would be subject to aerodynamic heating temperatures up to 2,500 degrees F, and so would be cooled by a stream of liquid nitrogen. [15, p.207] [58, p.37] NACA engineers provided the Q-ball conceptual design, and detailed design and fabrication of six units was done under contract by Northrop Aircraft Corporation. The Q-ball and the YAPS head were physically interchangeable, and initial X-15 flight testing up to about Mach 3 would be done with YAPS heads. [26, p.41]
The Inertial Platform Reference System
Because of the high speed, high altitude operations, conventional barometric flight instruments such as the altimeter and air speed indicator were practically useless in the X-15s, except in low-speed, low altitude flight such as landing conditions. An inertial reference platform fed by accelerometers and attitude gyros was going to be needed to compute speed, altitude and earth reference angles for guidance and control. NACA worked out the conceptual design for the platform, and drafted a specification used by the Air Force to contract with Sperry Gyroscope Company for detailed design and fabrication. It would be one of the first inertial reference systems ever built, and because it was before the era of small powerful digital computers, it would use analog computation to generate the parameters sent to the cockpit displays and instrumentation recorders. The computer would develop: airplane pitch, roll, and heading angles with respect to the earth, airplane forward speed, vertical speed, down-range speed, cross-range speed, and altitude. These parameters would be sent to various cockpit instruments including an indicator mounted in the center of the instrument panel that showed roll, pitch and heading angles on a pictorial “8-ball.” The 8-ball also showed angle of attack and yaw angle measured by the ball nose. These measurements were shown by long horizontal and vertical pointers, respectively, superimposed on the 8-ball. [58, p.43]
The inertial platform weighed about 165 pounds and occupied three cubic feet. It would be aligned from reference systems in the B-52 before launch, and had a limited period of accuracy just long enough for a typical X-15 flight. The demands placed on it were, in reality, exceeding the state of the art of gyroscope stabilized systems of the day. It would be in a constant state of rework and improvement during the X-15 program, all new components being designed and built at the Flight Research Center (FRC), formerly the High Speed Flight Station. Eventually the entire original system would be replaced by components built at the FRC, and would be redesignated the FRC-66 Analog Inertial System. [26, p.195]
The X-20 Dyna-Soar project of the late 1950s and early 1960s was to be the next manned research vehicle beyond the X-15, and was intended to put a rocket propelled space plane into orbit. Like the X-15, it would be able to glide to earth under pilot control, and land at an airport. It was, cancelled in late 1963, however. Minneapolis-Honeywell Regulator Co. had been contracted to develop a digital inertial guidance system for the X-20, and upon project cancellation three of these became available for X-15 project use. With digital computers and new technologies, these platforms would be a significant improvement over the FRC-66 analog system, and by 1965, the smaller, more accurate Honeywell systems were installed in all three X-15s. [26, p.195]
Engine Trouble
North American Aviation hosted an X-15 mockup inspection in December 1956, and at that time the Air Force reported that the XLR99 engine was eight months behind schedule. By February 1958 the engine was one year behind schedule, whereas the X-15 airframe was on schedule. [7, p.251, 292] In July the Air Force and NACA began considering an interim engine so that the start of flight testing would not be delayed. The testing program called for starting at the low end of the speed and altitude range and methodically working to higher speeds and altitudes, so an engine with lower thrust than the XLR99’s 57,000 pounds could be used to start the program. The project mangers decided that the tried and true Reaction Motors XLR11 engine that had powered the X-1 and D-558-II series would be the interim engine. In place of the large engine, two XLR11s could be fitted into the engine compartment, and it was found that the smaller engines could use the existing propellant tanks with minor piping changes. The XLR11 was not throttleable, but each of the four chambers could be fired separately so the two engines would provide eight levels of thrust. They were able to make up a dozen XLR11 engines from parts available at Edwards, and would install four in the first two X-15s. The remainder would be spares and also used for ground tests. The third X-15 would stay at the North American plant and wait for the XLR99 that was expected to be available in 1960. [7, p.251, 292, 295] NAA found that installing the interim engines was surprisingly easy. They were installed slightly canted so that their thrust lines passed through the plane’s center of gravity to prevent a rotating moment. The weight of the two XLR11s was slightly more than the XLR99 weight. [26, p.183]
Rollout and Delivery
In November 1956, North American froze X-15 design and proceeded to build a full scale mockup that NACA and the Air Force inspected, and passed with minor changes in December. This was six months after the date of contract, and within schedule. Construction of the three planes would take two more years after the design freeze, and “rollout” of X-15-1 complete with two XLR11 engines was on 15 October 1958, two weeks ahead of schedule. Vice President Richard Nixon, a California resident, was the principal speaker at the rollout. Structure weight was 325 pounds under the specified design weight, and even with the heavier XLR11 engines and the 700 additional pounds of instrumentation it was only 100 pounds heavier than original specified weight. [7, p.250, pp.302-303] [26, p.85] The number one plane was trucked to the NAA hangar at Edwards, covered by heavy brown wrapping paper, on 17 October 1958. High Speed Flight Station technicians immediately began instrumentation installation including the traditional NACA nose boom and YAPS head. This was then followed by five months of static ground testing of structure and systems. [52, p.81] X-15-2 would arrive at Edwards two weeks after No. 1. [7, p.304]
The David Clark Co. MC-2 Full Pressure Suit
The X-15 cockpits were pressurized, but if cabin pressure should fail at the altitudes expected of the X-15, the pilots would be killed almost immediately by boiling body fluids. A full pressure suit for the pilot was mandatory in case of emergencies. Partial pressure suits had been worn by military pilots for years, and even full pressure suits had been around for many years. For example, aviator Wiley Post had worn a full pressure suit during his 1935 coast-to-coast stratospheric flight attempt in his Lockheed Vega. The problem with existing full pressure suits was they more resembled a diving suit than aviator’s garb. They were bulky, heavy, and severely constrained the pilot when inflated. Their weight and bulk were out of the question for the X-15 project; a new full pressure suit was needed. Air Force Major Ralph Richardson and life support specialist Roger Barniki of NASA took on the task of managing the pressure suit development and acquisition. Richardson was in charge of the Edwards Air Force Base high altitude chamber, and had his own shop for building and repairing pressure suits including their mechanical parts. Former HSFS test pilot Scotty Crossfield would be the North American test pilot for phase I X-15 flight testing, and since he would be the first to be fitted with the new pressure suit, he took a great interest in its development. [7, pp.238-243]
The Air Force contracted with the David Clark Company of Worcester, Massachusetts, to develop the new suit. The company had manufactured partial pressure suits, and in 1951 was developing a full pressure suit for the Navy, however, the company’s main product line was not pressure suits, but rather women’s girdles and brassieres. Clark was originally a manufacturer of women’s wear, who began to specialize in undergarments. He was a tinkerer and inventor who before WW II had devised a knitting machine that could knit a single-piece girdle out of elastic thread. During World War II his production had been diverted from women’s wear to aviator’s anti g suits (partial pressure suits). Pressure suits and other aviator protective wear were sort of a challenging hobby for him, and he financed most of his own research and development. Full pressure suits had to be tailored to the frame of each pilot and Crossfield would make many visits to the Worcester plant where he would contribute significantly to the design, development, and testing of the suit, [7, pp.238-243] [15, pp.174-175]
A link-net material devised by David Clark was a key component of the suit and had been inspired when Clark took apart a woven Chinese finger trap. The link net would hold an airtight suit layer against the body regardless of movement or position. The 39-pound suit was made of five layers: 1. long winter underwear, 2. ventilation garment to cool the pilot, 3. rubberized airtight pressure garment with anti g bladders, 4. link net material, 5. outer protective coveralls. Originally, the outer coveralls were made of an olive drab material that greatly disappointed press photographers who though that a “space suit” should be much more photogenic. During a visit to the plant, Crossfield noted a bolt of women’s dress material that had been coated with an aluminum layer. It was called silver lame. Crossfield asked Clark if the outer garment could be made of the aluminized dress material. Clark made one, much to Crossfield’s approval, and they decided that the official reason for the new outer garment would be that it was a better reflector of heat. [7, pp.255- 261] In the original suit, the five layers would be put on one layer at a time, taking about 25 minutes. It would be cooled by nitrogen gas that would be kept out of the pilot’s helmet by a neck seal. Breathing oxygen would be fed into the helmet. The suit also included sensors that monitored pilot respiration, ambient & suit pressure, and electrocardiogram pickups; all of which would be telemetered to the ground while in flight. These recordings would be of much value to the future space program. [7, p.315] [52, p.82]
Users generally adjudged that the David Clark MC-2 full pressure suit was the first in the world that allowed the wearer acceptable mobility, even though when inflated it would be found to be slightly constraining and awkward. However, the suit would only inflate in emergencies when normal cabin pressurization was lost, and when uninflated was reasonably comfortable. The suit would eventually become the standard pressure suit for the Project Mercury astronauts and would serve as prototype for Moon Project Apollo pressure suits. [7, pp.238-243, 323] [25, p.43]
Improvements would be made to the MC-2 suit as the X-15 project progressed. By 1961 an improved David Clark suit designated A/P22S-2 would be available. This suit integrated all suit layers, except the outer aluminized protective garment, into one suit. The redesign was made possible by a new zipper that could provide a pressure seal up the the back of the suit. The neck seal would be eliminated in favor of a seal around the face that was more comfortable. First flight use would be by NASA pilot Joe Walker on 30 March 1961 in X-15-2. The MC-2 suit would be used on 36 X-15 flights, and all other flights would use the A/P22S-2 suit. This suit would evolve into the standard U.S. high altitude pressure suit, including space shuttle use.[26, p.207]
The High Range
The rocket powered research airplanes preceding the X-15s had normally been released from 30 to 40 miles from Muroc Dry Lake, and the instrumentation radar at the High Speed Flight Station had no trouble tracking them over their complete flight path. Communication radios and radio telemetry similarly had no problem. The X-15 flights, however, were going to span a distance of around 400 miles, so they were going to need an instrumented range with more tracking sites. In May 1959 the Air Force and NACA entered an agreement whereby they would jointly work out X-15 testing range requirements and plan the range, after which the Air Force would build and equip the facilities. When built, NACA would operate the range that would officially be called the High Altitude Continuous Tracking Range, but was usually called the “High Range.” Gerald Truszinski, chief of the HSFS Instrumentation Division, took on the job of range layout, general location of facilities, preliminary design and preparation of a contract work specification. By November 1955 he had determined that the range would have three tracking sites, with the High Speed Flight Station serving as the terminal site. The range would be laid along a rout having a number of dry lake beds for emergency landing purposes, and the best series of dry lake beds seemed to be on a line running from Wendover, Utah, to Edwards. The exact location of the two field sites would be determined by the range contractor. [52, pp.80-81]
In March 1956 the Air Force awarded a contract to Electronic Engineering Company of Los Angeles to design, construct and equip the sites. [26, pp.52-53] Construction started in 1956 and was complete by July 1959. Final range dimensions were 484 miles long, with 50 mile wide flight corridor. [52, pp.80-81] The two remote sites, located near Beatty and Ely, Nevada, were built on hill tops, and new access roads had to be built in to them. They were so remote that they needed their own electrical generating plants, and neither site could sustain a water well, so water had to be trucked in to fill storage tanks. [26, p.54] Some of the crew members would travel to the sites before each flight, but some crew was permanently stationed at the sites and lived in nearby towns. [58, pp.58-59] Each of the three sites would be equipped with precision FPS-16 radar sets that had been declared surplus from ballistic missile tracking sites. The radar data would be used to calculate aircraft position, airspeed, altitude, and reentry location prediction; all with analog computation. The calculated geographic and altitude information would be plotted on pen chart recorders to show airplane location and flight profile. Each site also had optical film recording oscillographs and magnetic tape recorders to capture the 87 channels of data telemetered from the X-15 to ground. Telemetered data such as airplane speed and altitude, would also be displayed for the flight controllers at each station. [52, pp.80-81]
Probably the most critical phase of a drop flight was engine light-off. If the pilot could not get an engine light after two attempts, he would have to make an emergency landing on one of the dry lake beds. If the X-15 were up to maximum speed or maximum altitude, it could glide for about 400 miles, but engine failure at launch point meant the pilot had very little gliding distance in a start from about Mach 0.8 and 30,000 feet. [58, p.53, 55] This meant that each drop had to be made near an emergency dry lake bed, called the “launch lake.” There had to be a number of launch lakes to accommodate the different kinds of flights. For example, a speed flight in the lower atmosphere needed a range of over 400 miles, whereas a high altitude attempt would need about 300 miles, and an entry heating test about 200 miles. Six dry lake beds were initially selected as launch sites, and if the X-15 had to land on one of these it would have to be transported back to Edwards on a flat bed truck. About 15 lake beds were also picked as intermediate emergency landing sites in the event of mid-flight emergency. The X-15 landing slide-out could be as much as two miles which dictated the minimum size of the landing lakes. The Air Force marked the landing areas with tar strips 8 feet wide and 300 feet apart. The standard 300 foot separation was used to facilitate pilots height perception, because the flat, featureless lake beds gave very little clues for judging altitude. [58, p.50] By the end of the program two more dry lake beds would be used as launch lakes. [26, p.57] The Mojave Desert winter featured a rainy season that was both good and bad. The good part was, wind-blown water would be pushed back and forth over the lake beds, smoothing them and healing imperfections, but the bad part was it prevented winter flights and took time for the beds do dry out in the spring. A part of the annual spring ritual included a trip to each lake to measure if the bed was firm enough to support a landing. The testing tool was a six-inch diameter lead ball dropped from an exact five feet. If the diameter of the resulting impression was less than 2.5 inches, the lake was adjudged OK for landing. [58, p.51]
Central control for the drop flights would be in the Flight Research Center control room (NASA-1), and one of the X-15 pilots would serve as primary mission controller. From his displays of X-15 geographic flight path, flight profile, speed, and altitude, the controller could advise the pilot to climb, dive, turn, etc. to stay on the planned flight path. Also, the night before an X-15 flight, selected X-15 pilots would fly to the two remote tracking sites to act as controllers in the event the X-15 might have to be directed to a landing a a nearby emergency landing lake. [58, p.59, 163,167] During X-15 flight testing, the pilot’s monitored heart rate would be telemetered to the ground and recorded, and it was found, curiously, the the highest recorded rates were during the pre-launch checks when the pilot was anticipating drop. Once dropped, heart rates would generally lower as the pilot concentrated on flying. X-15 pilot Milt thompson notes in his book that when the flight controllers were monitored by an electrocardiogram machine, their heart rate was generally faster than the pilot’s elevated heart rate. [58, p.248]
Simulator and Centrifuge Testing
The high Speed Flight Station had started an analog computer flight simulation program with the X-2, and the X-15 project would see flight simulation brought to a new level. The Air Force contract with North American Aviation included a ground based analog computer flight simulator that included a realistic X-15 cockpit, and the complete X-15 flight control system including all cables, push rods, bell cranks, hydraulics, and control surface masses. To accurately replicate the plane’s internal control linkages, the “Iron Bird” was almost as long as the X-15 itself. It also included the three-axis stability augmentation damping system, and it could simulate flight from Mach 0.2 to seven and altitudes up to 200 miles. The simulator would initially be based at the North American plant until their phase of X-15 flight testing was completed. Then the simulator, less the three Electronics Associates PACE 231R analog computers, was to be delivered to the High Speed Flight Station in 1961 where it would be powered by the Station’s Electronics Associates analog computers (technically called differential analyzers) .[26, p.45] Initial airplane flight parameters came from wind tunnel model testing, and as wind tunnel data was verified or modified by flight testing, the updated airplane description parameters and stability derivatives would be entered into the analog computer. X-15 pilot Milt Thompson states that he considered the flight simulation absolutely essential to the project, and he could not have successfully made his flights without the simulator practice. [58, p.70]
In the late 1950s, analog computers were king when it came to flight simulation because digital computers of the day lacked the computational speed to simulate even two degrees of freedom in real time. This writer was occasionally detailed to fly the simulator, that we called the Iron Bird, in the wee hours of the morning at the NAA plant to generate recordings of control system interactions in various flight profiles. I can vouch that the device was extremely realistic except for the lack of feeling actual flight accelerations, and lack of a realistic landing display. I often ended up many feet under ground, or stalling out too high on landing, with equally disastrous results. Fortunately the whole thing could be reset with the push of a button. To generate some kind of a feeling of the accelerations involved in actual X-15 flight, I half jokingly proposed a system of mechanized seat straps that would pull the pilot down and/or back into the seat to simulate longitudinal and normal accelerations, and inflatable air bladders behind the occupant that would push forward to simulate high deceleration. But Scotty Crossfield had a better idea.
The Aero-Medical Acceleration Laboratory of the Naval Air Development Center at Johnsville, Pennsylvania, had a human centrifuge capable of generating the magnitude of acceleration forces expected in X-15 flights, including rocket thrust on climb out, reentry deceleration, high g pullouts, and other normal and abnormal flight maneuvers such as in-flight dynamic instability. The machine had a fairly capacious gondola that could be fitted out to simulate a cockpit, and the gondola was mounted in gimbals so it could be oriented in any direction while whirling in a circle to simulate accelerations in any direction. In particular NACA and North American were interested in how effectively a pilot would be able to use the X-15s right hand wrist actuated flight controller in conditions of high g loading, including violent maneuvers of the kind experienced by test pilots Chuck Yeager and Milburn Apt in their encounters with roll coupling. Accepting Crossfield’s proposal, NACA not only was able to get considerable time on the centrifuge, but the Navy helped modify their system to allow the pilot in the gondola to realistically simulate dynamic flight. [56, Chapter 4]
Initially, the Navy had rigged the centrifuge to simulate flight g-force profiles in a programmed manner with precut cams, but what the X-15 project needed was a way of varying the acceleration forces to represent airplane dynamic responses to pilot control actions. What this took was interposing an analog computer simulation of X-15 flight dynamics, that took pilot inputs from a replica X-15 cockpit in the gondola, and in turn computed control orders to vary centrifuge speed and gondola orientation. The mechanization was considerably more difficult than the Iron Bird simulator, but it worked. X-15 pilots were dispatched to Johnsville and made about 400 simulated flights long before the first real X-15 flight. Many of these flights simulated reentry's with failed systems, or unusually high angles of attack or yaw angles. In most cases the accelerations were more severe than actually experienced in later X-15 flights, and the testers found that the side located wrist controller worked well under the worst of conditions, in fact up to 12 gs. The centrifuge flights also made useful contributions to improvements of the side controller, the pilot’s restraint system, and instrument displays.[56, Chapter 4]
