First-Hand:The X-15 Project - Flight Testing - Chapter 12 of the Experimental Research Airplanes and the Sound Barrier

By David L. Boslaugh, CAPT USN, Retired

Research Goals

The X-15 program’s reason for being was not to set records; they would be incidental to the real purpose of developing and providing information on technical aspects of hypersonic flight. Also it was desirable to at least gain a toe hold of knowledge about flight out of the atmosphere. NACA/NASA would then disseminate new knowledge and lessons learned in a series of reports, some classified and for the armed services primarily, and others intended for all industry and academic institutions. Areas of research would include:

• Determining numerous airplane aerodynamic coefficients such as lift and drag coefficients at speeds and altitudes over the flight envelope to compare with wind tunnel results for validation of tunnel designs and testing procedures.
• Determining static stability derivatives such as pitch and yaw restoring moments for same purpose as above.
• Determining dynamic stability derivatives such as oscillation damping ratios about all three airplane axes for same purpose as above.
• Validation of various analytical tools used to predict airplane behavior.
• Measuring air pressure distributions at hundreds of locations on the airframe.
• Measuring structural loads experienced in hypersonic flight.
• Measuring aerodynamic heating effects experienced in hypersonic flight.
• Measuring a number of pilot biomedical parameters, especially the effects of high accelerations and weightless flight.
• Evaluating various automatic control system and stability augmentation system designs.
• In the later stages of the program, the X-15s would serve as platforms to carry numerous non-X-15 related experiments to high speeds and altitudes.

The premature loss of the X-2 research airplane in September 1956 denied the X-15 project needed information on airplane stability and control in the region from Mach 2.5 to above Mach 3, so wind tunnel operations and techniques at these speeds had not yet been validated by actual flight experience. Flight investigation and validation in this region was was going to be another X-15 chore. [7, p.247]

Preparing for a Test Flight

There would be three kinds of flights: speed, altitude, and heating. Of these, the heating flights demanded the most piloting precision, especially in holding a precise angle of attack. [15, p.239] These flights would be done at altitudes around 80-000 to 90,000 feet, at high speeds, and usually continued until engine burnout. The pilot would start a gentle pushover at about 70,000 feet for a heating flight so that he could level out at around 100,000 feet or less, and all control would be done with aerodynamic control surfaces. [58, p.189] For altitude flights, the pilots continued upward at a fairly steep angle and either continued to engine burnout or shut the engine down at a predetermined speed or altitude. They would spend two-to-five minutes out of the atmosphere in a weightless condition. [26, p.92] When above 200,000 feet, the aerodynamic controls had no effect, and pilots had to use the reaction control system for attitude control. [7, p.198] Speed flights would be done at the lowest altitudes, around 50,000 feet, and would have the longest ranges, around 400 miles. Control would be with aerodynamic control surfaces only. For speed flights, it was NACA practice to increase speed by one half Mach number for each successive flight. [15, p.322]

In preparation for each flight, a flight planning engineer would consult with specialists in areas such as stability and control, structures, and heating to determine their reaserch requirements and then work up the sequence of test maneuvers requested of the pilot. The flight planners were senior Flight Research Center engineers who not only were experts on the X-15, but also had worked on flight planning on previous research airplanes for years. They were general experts on research airplane performance, stability and control and had published research reports many times over. [26, p.87] The flight planner would work out the distance to be covered in the test, and would then select the launch lake and emergency landing lakes. They would then would practice the flight in the simulator numerous times before working with the pilot for a number of days coaching him in simulator runs. Once the pilot had memorized the flight profile, the flight planner would simulate emergencies for the pilot. Pilots would spend up to 20 hours in the simulator preparing for a specific flight, and a single mission might be re-flown as many as 200 times. [58, pp.68-69] In addition to the simulator runs, before each flight the pilots would practice landings in an F-104 fighter with flaps down and in engine idle, which closely simulated the landing characteristics of the X-15. The pilot would usually make 75 to 100 landings in the F-104 before each X-15 flight. [15, p.117]

The Mother Planes and Chase Planes

The initial X-15 contract with North American called for the Air Force to provide a ten engined (six reciprocating, two jet) B-36 intercontinental bomber as the X-15 mother airplane. The contract called for NAA to make the modifications necessary to turn the bomber into a research craft launching platform. The huge craft had three bomb bays that would be modified to carry the research airplane half submerged into the lower fuselage. The B-36 also conferred the advantage of allowing the X-15 pilot to ride aloft in safety and comfort inside the B-36. North American even planned to build a special heated compartment just ahead of the X-15 boarding catwalk to accommodate the research pilot and two helpers. There were some disadvantages with the B-36 however. The bomber could only achieve a Mach 0.6 launch speed that would detract from X-15 performance, furthermore, the B-36s were getting old. The Air Force grew increasingly concerned that the B-36 would become a maintenance problem because the planes were being phased out of inventory. Thus, a search for a different launching platform.

After ruling out the Convair B-58 “Hustler” delta winged supersonic strategic bomber and the KC-135 tanker, the Air Force tasked North American, in May 1957, to study the feasibility of using a B-52 strategic bomber as the mother craft. The study found no particular problem, other than the X-15 could not be carried under the fuselage because of the bomber’s landing gear configuration. Instead, the X-15 pilot would have to enter the X-15 cockpit on the ground and ride aloft suspended under the B-52’s wing inboard of the two inner turbojet engines. There were also some distinct advantages: the B-52 could launch at Mach 0.8, and sister airplanes were expected to be in inventory for a long time to come. In fact, they are still on active duty in 2016, almost 60 years later. In June 1957, the project mangers decided to switch to the B-52. [26, pp.65-67] The Air Force found two very old B-52s, the third and eighth B-52s built, and the NAA contract was amended to modify the two bombers as mother airplanes. [7, p.266] It can be recalled that two X-1 series pilots lives had probably been saved because they could get out of the X-1 cockpit into the mother plane before their craft had to be jettisoned after in-flight explosions. Therefore the decision to carry the X-15 under the bomber’s wing was not taken lightly. One saving grace was the ability of the X-15 pilot to eject while being carried, because the cockpit would be ahead of the B-52s wing. But, this meant the research pilot would be confined in the X-15 cockpit for about one and one half hours before launch.

When suspended under the bomber’s wing, the X-15s tail was going be in the way of the B-52s wing flaps so the flaps were going to have to be deactivated. The fueled X-15 would weigh 32,215 pounds, whereas the B-52 could carry a 70,000 pound bomb load, so the X-15 weight was not a particular problem. However taking off and landing without flaps was going to pose a challenge to the first mother plane pilots, Air Force Captains Charles Bock, and John E. Allavie. They would have to devise and test new landing and takeoff procedures, bearing in mind that the B-52 would occasionally have to make landings with the research airplane still attached. [7, pp.264-265]. The X-15’s liquid oxygen tank contained 1,038 gallons, of which as much as 800 gallons could boil off during the climb to launch position. To compensate, the B-52 was fitted with a 1,200 gallon liquid oxygen tank and automatic oxygen transfer piping connected to the X-15. From 600 to 800 gallons would be transferred in a normal flight. Other modifications to the two mother planes included a wing pylon with shackles for holding the research airplane, turning the former electronic countermeasures compartment into a launch panel operator’s station, provision of piping to supply the X-15 pilot with breathing oxygen, cabling to provide electrical power to the X-15, and cabling to provide alignment information to the inertial platform until launch. [58, p.162] In addition to the B-52 pilot, copilot, and launch panel operator, the fourth crew member would be the flight engineer. These four would have ejection seats, and NACA/USAF policy discouraged any additional “passengers” because they would not have an ejection seat. Twenty-four USAF pilots would eventually fly the two B-52s. Of them, Major Fitzhgh L. Fulton would make 94 mother-ship flights. [15, p.18]

Chase planes would be essential to support the research flights to: monitor the X-15 while hanging from the B-52, monitor as much of the powered flight as possible, help coach the X-15 pilot to an emergency landing site, and coach the pilot for a normal landing on Muroc Dry Lake. A minimum of three and a maximum of five chase planes would be used on all X-15 flights. These were designated:

• Chase 1 - prelaunch chase, would stay with the B-52 (Usually a USAF F-100)
• Chase 2 - launch chase, would try to keep up with X-15 upon launch, and stay with X-15 if it had to make an emergency landing. (An NACA F-104)
• Chase 3 - intermediate chase, covering emergency landings at intermediate emergency landing sites. (An NACA F-104)
• Chase 4. - Edwards landing assistance. (A USAF F-104)
• Chase 5. - Second intermediate chase for long high altitude flights. (An NACA F-104)

[58, pp.60-62]

First Captive Flights

North American Aviation would be responsible for the Phase I contract demonstration flights that called for demonstrating speed up to Mach 2 and altitudes up to 100,000 feet. Most contractors supplying aircraft to the Air Force leased property along the dry lake bed at Edwards, and had their own flight testing hangars. The NAA hangar at Edwards was about twice the size of the High Speed Flight Station facilities facilities, and was home to about 100 engineers and maintenance personnel. They even had their own control tower, and during contractor flight testing the X-15s would be based at the NAA hangar. [7, p.299, 330, 391] The NAA hangar was not far from the HSFS, and the Air Force selected a location near the two facilities to set up a “mating area” for loading the X-15 onto the B-52. Here the Air Force installed two hydraulic lifts flush with the apron. The X-15 would be positioned with the nose wheel on the front lift and the two landing skids on the rear lift. The plane would be raised to just below the wing pylon so electrical and piping connections could be made, and then lifted for connection to the three release shackles in the pylon. The X-15 landing gear would then be manually retracted. [26, p.104]

X-15-1, equipped with two XLR11 rocket engines, arrived at the NAA hangar on 17 October 1958 to start five months of static load testing during which NACA technicians installed its instrumentation including their traditional nose boom with YAPS head. By March, it was ready for captive flights followed by unpowered glide flight. [52, p.81] The second X-15 arrived at Edwards two weeks after No. 1. Scotty Crossfield went aloft strapped in the X-15-1 suspended under the B-52 for the first captive flight on 10 March 1959. Purpose of this flight was to check out: aerodynamic flight control movements, landing gear deployment, liquid oxygen top off system, cabin pressurization, pressure suit inflation, cockpit instruments, NACA instrumentation, telemetering system, and communications radios. None of this was accomplished, however, due to the sudden appearance of dense smoke in the cockpit. Upon landing, it was found that the generator on one of the two auxiliary power units (APU) had seized and burnt up. It would take two weeks to get the craft ready for the second flight attempt on 1 April. On this flight, the APUs worked, but a number of problems arose in Crossfield’s full pressure suit, and the communications radio failed. The X-15/B-52 combination went aloft for the third time on 14 April, but this time the number 2 APU froze up and tore off its mounts. This triggered an intensive APU “get well” program that would eventually pay dividends. All went well in the fourth captive flight on 21 May, that is until the nitrogen-cooled number 1 APU bearings began to overheat, and in a mistaken switch movement, Crossfield shut down number 2 APU instead of number 1, quickly followed by the seizure of the number 1 unit. After APU replacement, Crossfield and NAA managers decided that if all went well in the next captive flight, Crossfield would launch for the first glide flight. [7, pp.300-337]

First Glide Flight

The Sidearm Controller

The prototype X-15 sidearm controller installed in the HSFS F-107 fighter for flight evaluation. All of the Station’s test pilots had a hand in designing and evaluating the controller. The knob at the top of the handle is for adjusting pitch trim. NACA photo

The X-15s were expected to undergo high accelerations along all axes, for example the rocket engine at full thrust could produce five times the force of gravity pushing the pilot back in his seat, at burnout the craft could suddenly decelerate at up to four gs due to aerodynamic drag, and maneuvers could produce pilot up/down accelerations up to 7 gs. There was concern that such high accelerations could keep the pilot from effectively using a conventional center stick, and the proposed solution was a wrist activated right-side located controller that could be used under high accelerations. We called it the sidearm controller, and it was rigged to move the center stick by a mechanical linkage connected to the bottom of the center stick. The pilot would not have the same strength and leverage when manipulating the side controller with wrist actions as he would in moving the center stick, so a boost, much like power steering in a car, was provided by small hydraulic cylinders connected to the side controller. The system was arranged so that it would amplify the pilot’s force, and at the same time give him feedback feel of stick force.

Using the F-107 fighter and a spring-mounted pivoting slide projector in a fairly accurate replication of the X-15’s pitch dynamics while evaluating the X-15s side located controller. Pitch angle was projected as a line on the screen, and pitching motions were also recorded by the oscillograph recorder next to the slide projector. The Station’s DC-3 utility airplane, in which Marian occasionally flew as “copilot” with Neil Armstrong, is in the background. NACA photo

High Speed Flight Station engineers and test pilots designed and tested the prototype side controller, including contributions by Scotty Crossfield and Neil Armstrong. Surprisingly, they found that human wrist mechanisms could be significantly different from one person to another. What was comfortable to one pilot was not to another, but they finally found a design compromise in controller articulation acceptable to all the test pilots. [20, p.153] High Speed Flight Station engineers and technicians built the prototype controller and installed it in the Station’s F-107 fighter, my project airplane, for flight evaluation. One of the purposes of the flight evaluation was to determine the best boost ratios of the small hydraulic actuators that gave the pilot his “power steering” assist. The boost ratios could be adjusted by resetting the two pressure regulators that supplied hydraulic pressure to the boost actuators. Before X-15 flight testing, we ran a simulation of the X-15 longitudinal dynamics with the side controller and pilot in the loop to get a feel for what the boost ratios should be. In this case we did not use the station’s analog computer flight simulator, but rather concocted a simple, but fairly accurate, analog representation of the X-15 dynamics. The F-107s internal hydraulic control system was quite similar to the X-15’s system, and could be made to closely replicate the research plane’s system by adjusting internal gains. To simulate X-15 longitudinal flight dynamics we mounted a fairly heavy commercial slide projector on a pivot, and connected the back end of the projector to the F-107s horizontal stabilizer with a coil spring. We had a number of springs that could make the slide projector bounce in pitch at the X-15s natural longitudinal frequency at a number of speed and altitude combinations. A magnetic damper was connected to the projector to simulate the plane’s natural aerodynamic pitch damping. It also could be adjusted to replicate the research plane’s damping ratios at a number of speed and altitude combinations.

Scotty Crossfield’s Wild Ride

The time for the X-15’s first glide flight was drawing near, but we sill had some simulator runs to make before deciding on the best boost ratios for the side controller. The side controller was intended for conditions of high acceleration and would not be needed in a glide flight so I told my boss Jack Fischel that the North American Engineers should tell Scotty not to use the side controller because the gains in the pressure regulators were not properly set yet. No big issue was made of it because there was no need to use the side controller on a glide flight, and the center stick offered much more precise control for a maiden flight. Things went well on the 8 June 1959 captive flight, and, as planned, Scotty launched. Crossfield later wrote in his book “Always Another Dawn”, “My right hand moved to the sidearm control handle, which I had elected to use on this flight.” He had been much involved in the design of the controller, and was itching to try it out. He then notes that almost immediately after launch the plane’s nose unexpectedly pitched up, and he quickly applied nose down control with the side controller. Instead of just leveling off, the nose pitched down, prompting him to apply a nose up control that overshot the mark even more. Subsequent attempts at trying to level off resulted in a pitching oscillation that was getting worse as he was getting closer to the lake bed.

This is X-15-2 in a fairly normal landing. The X-15-1’s nose was much higher at landing impact of Scotty Crossfield’s first glide flight. NACA photo

The staff at the High Speed Flight Station had been told that Scotty would probably make the first glide flight that day, so Marian and I found a convenient spot on the roof to watch the landing. Technicians had even set up speakers on the roof so we could hear the radio talk. We could see a black speck approaching the far side of the lake bed, and I turned away momentarily to se where the mother plane was because it was going to make a low pass over the station which would also be pretty spectacular. As I turned away, I heard Marian exclaim with great alarm, “What’s he doing?” I turned back to see the X-15 in a wild pitching oscillation, and settling fast toward the lake bed. In quite a feat of piloting skill Scotty managed to impact the lake with nose high in the air. If it had been on a nose down swing, it would have been all over for Scotty and X-15-1.

Marian the librarian and an Air Force F-104. Photo by the author

Jack Fischel told me to get the instrumentation film developed as soon as possible, and try to figure out what had happened. It did not take much analysis to see that there was an abnormally long time lag between control stick motion and elevator response. It was not characteristic of the well tuned X-15 main control system, but rather looked like the sidearm controller whose boost actuator gains had not yet been set. I told Jack, “I think Scotty was using the sidearm controller, and it was a pilot induced oscillation.” Jack responded, “But they were supposed to tell him not to use it.” He picked up the phone, called the North American hangar and asked for Scotty Crossfield. He put the question,”Scotty, did you use the side controller.” A pause. “But, didn’t they tell you not to use it?” Jack’s face was ashen when he put the phone down. “I don’t think anyone told him.” One good thing about Scotty’s use of the sidearm controller was the instrumentation recordings that gave valuable information on the correct boost gain settings, and North American engineers soon had the pressure regulators set, with the end result that there would never again be a pilot induced oscillation caused by the side controller. NACA test pilot Milton Thompson would later write in his book “At the Edge of Space” that most X-15 pilots came to prefer using the sidearm controller in all flight conditions, including landings. He noted that it became a “macho thing” for him to use the side controller for the entire flight even though there were times when he could have gotten more precise control with the center stick. [58, p.42] Scotty’s hard landing permanently sprung the X-15’s rear skids and caused other structural damage severe enough that the number one plane had to be returned to the North American plant for six months of repairs. This left X-15-2 to be readied for the first powered flight. [17, p.240]

Farewell to the High Speed Flight Station

The University of Southern California gave courses at the Edwards AFB Test Pilot’s School that could lead to a master’s degree in aeronautical engineering, and I had been taking one evening course per semester since soon after starting at the High Speed Flight Station. Most of the students were either test pilots or flight test engineers, which made the courses doubly interesting because of their experiences and knowledge. Some times a theoretical question would come up in class that could not be answered, and one of the students would volunteer to work it into his next flight and report back. My problem was, it was going to take a long time at one course per semester to qualify for a masters degree, but I found another solution. Commander Fernandez was the Navy Liaison Officer at the Air Force Flight Test Center, and in addition to inviting me to watch demonstrations of various navy projects, he passed me all the Bureau of Naval Personnel (BUPERS) instructions and notices that were routinely sent to him. One of these, in the summer of 1958, was of special interest; it described the application process for the U.S. Naval Postgraduate School at Monterey, California, and it listed aeronautical engineering among the curricula. Marian and I had a discussion about staying in the Navy as a career if I were accepted for postgraduate school, and in September I mailed my application.

In January 1959, BUPERS responded that I had been selected, but not for my first choice of aero engineering but rather for second choice: electronics engineering; in particular the Information and Control Systems curriculum. I had picked the latter curriculum as second choice because it had a strong content of servomechanisms, my specialty at the Station. Another part of the curriculum was something new called digital computers, that I equated with things like accounting or maybe data reduction, but not with automatic control systems or flight dynamics. I would learn later that, at that time, the Navy was making its initial foray into using digital computers at sea in weapon systems, and that it had a rapidly growing need for engineers educated in the new technology. Thus the assignment to the second choice curriculum.

Marian and I departed the High Speed Flight Station for Monterey, California, In July of 1959, and by this time she had contracted the dreaded Library Disease. Before we left, Station management had offered me the job of X-15 project control system engineer if I would stay on as a civilian. I have often wondered what it would have been like had we stayed. The reader will see from the following narrative, that it would have been a very interesting, exciting, and challenging job.

First Powered Flight

X-15-1 in one of its first powered drops. NASA photo

On 24 July 1959, X-15-2 was carried aloft under the B-52, this for the first time loaded with nine tons of propellants. Purpose of the flight was to test the propellant jettison system, propellant tank pressure regulators, pressure suit inflation, and liquid oxygen top-off from the B-52. Everything worked except the oxygen top-off. The frozen-closed oxygen line was corrected, and the next flight was to be powered. The attempted powered flight on 4 September was not to be, however, because of a failed helium pressure regulator to the liquid oxygen tank. [7, pp.352-353, 356] Finally, on 17 September it all came together. All systems worked, and Crossfield fired all eight rocket engines. Alternating between side controller and center stick, Crossfield applied pulse and step motions over a range of speeds and altitudes. He experienced very little transonic buffet as he passed the “sound barrier” and leveled off at 50,000 ft. for a speed run. He found the craft responsive to the controls up to past Mach 2, the planned maximum speed, and shut off three rocket barrels at Mach 2.3. While decelerating, the remaining five barrels consumed the remaining propellants at 230 seconds burn time. During his landing approach, just as he blew off the lower ventral fin, the main propulsion alcohol pump casing broke and hot engine parts ignited leaking alcohol, causing considerable engine compartment damage. [7, pp.359-366] Repairs took until mid October.

On 27 September 1959, the High Speed Flight Station was renamed the NASA Flight Research Center. This meant that the new Center was no longer an adjunct of the Langley Laboratory, but had equal status with the other NASA labs and reported directly to NASA Headquarters in Washington. After two aborted flight attempts, Crossfield was again in the X-15-2 cockpit hanging under the B-52 on 10 October, his 11th time aloft and his third drop. Again he fired all eight barrels, and in spite of a failed roll damper, achieved a little over Mach 2, gathering more contract qualification data. [39, p.122]

Bent, But Not Broken

Crossfield’s fourth drop flight, and 13th time carried aloft, came on 5 November. After release, he got good ignition in the first six rocket barrels, but felt a jarring explosion when switching the final two. Almost immediately, the chase pilot reported fire in the engine compartment, and the fire warning light lit up. He had no choice but to shut all barrels down and begin propellant jettison. He was near Rosamond Dry Lake, and elected to land there with 1,000 pounds of non-jettisoned propellants still on board. The rear landing skids dug into the lake bed, and the extra propellant weight brought the nose wheel down hard. Crossfield was nonplussed as the craft slowed down much quicker than usual, taking only about 1,500 feet instead of the usual mile or more. The Air Force flight surgeon was the first to disembark from a helicopter and run up to the cockpit. He gave the good news that the fire was out, but the bad news was X-15-2 was busted in two. The belly dragging on the lake bed had caused the rapid deceleration. Investigation showed the explosion had been caused by igniter failure, and it was further found that the hard nose gear impact that caused the broken fuselage was due to foaming of hydraulic oil in the landing gear shock strut. It was found that the oil foamed every time the gear was released due to over-fast extension of the compressed shock strut. It had probably happened every time the plane landed, but this time the extra propellant weight had been the straw that broke the X-15’s back. The solution was to lower the gear sooner to allow the oil to settle. In the amazingly short time of 30 days, number 2 was returned to the NAA plant, the fuselage repaired and strengthened, the engine compartment rebuilt, and two replacement engines installed. In the mean time, X-15-1 was back at Edwards, and resumed the testing program. [7, p.381-387]

November 5th 1959, the result of Scotty Crossfield’s emergency landing on Rosamond Dry Lake. Despite the appearance of the damage, X-15-2 was repaired in 30 days and was flying again two months after that. NASA photo

By December 1959, X-15-1 was repaired and back at the Edwards North American Hangar. On the 15th Crossfield rode number 1 aloft, but had to abort the planned powered flight. On 23 January, he released from the B-52 with the stable reference platform installed for the first time. The platform worked well and he achieved Mach 2.6 and an altitude of 66,844 feet. [26, p.1] X-15-1 had completed its contractor qualifications, and NASA accepted the craft on 3 February 1960; five months behind schedule. A day later, Crossfield made his first flight attempt in the repaired number two plane, but had to abort. Eleven February saw a successful contract demonstration flight in X-15-2 when Scotty lit all eight XLR-11 chambers to make Mach 2.5 and climb to 90,000 feet followed by a dive where he demonstrated a high g pullout. [7, pp.390-391] On 25 March 1960, while Crossfield was making further contract demonstration flights in the number 2 craft, NASA pilot Joe Walker made the first government flight in number 1, achieving Mach 2 and 48,630 feet. [52, p.82] Crossfield would make two more structural load demonstration tests of number 2 at the end of March, including a 6 g pull up. Then in April 1960, North American took X-15-2 off line for a month to install the reaction control system. [7, p.393]

On 13 April 1960, Captain Robert M. White became the first Air Force pilot to fly an X-15 (No.1), in a familiarization flight achieving Mach 1.94. As of this date there were six government pilots assigned to the X-15 project: NASA pilots Joe Walker, Neil Armstrong, & Jack McKay; USAF pilots CAPT Robert White, & MAJ Robert A Rushworth; and Navy pilot LCDR Forrest S. Peterson. [52, p.82]. Joe Walker became the first to achieve Mach 3 in an X-15 on 12 May when he went to Mach 3.19 at 77,862 feet, then on the 19th, CAPT White became the first X-15 pilot to rise to over 100,000 feet; achieving 108,997 feet. [26, p.102] By mid May, the reaction control system had been installed in X-15-2, and Scotty Crossfield made the last contractor qualification flight of number 2, and also the last flight of that plane with XLR11 engines. During this flight he made the first test of the reaction control system, and for the first time since his first glide flight, he landed using the sidearm controller. He found the side controller a little sensitive, but no real problem. [7, p.396]

On 4 August, Joe Walker, flying X-15-1 beat the X-2 speed record of 2,094 mph by flying at 2,196 mph. [7, p.393] Then on 12 August, CAPT White exceeded another X-2 record:126,200 feet set by CAPT Ivan Kincheloe to achieve 136,200 feet; a new unofficial world’s altitude record. This would be the highest altitude achieved with the XLR11 engines. White would make a new unofficial speed record on 7 February 1961 by taking the X-15-1 to Mach 3.50. This would be the fastest flight using XLR11 engines; they had expanded the envelope as far as they could with the XLR11s. [15, p.78] [26, p.102]

The XLR99 Engine Arrives

The Reaction Motors, Inc. XLR99 rocket engine, capable of 57,00 pounds of thrust. It could be throttled from 30% to full thrust by varying the speed of the propellant pump, and could be restarted a number of times in flight. US Air Force photo

On 28 March 1960, Reaction Motors, Inc. delivered the first XLR99 engine; having about four times the thrust of the two XLR11 engines, but a year and a half behind schedule. This engine was to be installed in X-15-3 which had been held at the North American plant awaiting the engine. After installation there was to be a series of captive ground tests at the Edwards AFB rocket engine test facility. [7, p.399] On 8 June, number three was fastened down at the test facility with custom made steel clamps, and Scotty Crossfield was in the cockpit to run the engine controls for the first firing test. All other test personnel were in protective blockhouses. Crossfield started the engine at half thrust and then went to full power, after which he stopped the engine. He was then supposed demonstrate engine restart, and after waiting the required 25 seconds, he pressed the engine reset button. Simultaneous with the button press, he felt a violent explosion behind him, and the cockpit was suddenly accelerated forward by about 20 feet. Instrumentation later showed that the acceleration had been 50gs. Crossfield was helped from the cockpit unhurt, and looked back to see the state of the number 3 plane. What was left of the after half of the plane was still in the steel clamp, but the engine compartment was a shambles, and the ammonia and hydrogen peroxide tanks were in tatters. [7, pp.404-406]

X-15-3 after its explosion on the rocket engine test stand. What is left of the aft section can be seen still confined in the steel retaining clamp. The XLR99 engine was found not to be the culprit, however it was destroyed. NASA photo

Investigators determined that the cause of the explosion was an over pressured ammonia tank that had ruptured due to a malfunctioning pressure regulator. It was further found that the excess pressure had been caused by a feature of the test site, specifically piping used to route overflow ammonia to a diluting pond that had generated back pressure which the pressure regulator had failed to alleviate. The rupturing ammonia tank had, in turn, torn open the main hydrogen peroxide tank leading to an explosion of the mixed ammonia and peroxide. The new XLR99 had not been the culprit but explosion and fire had destroyed it. [26, p.108] All of the plane aft of the wings, amounting to about 65% of the craft was ruined but the decision was made to rebuild X-15-3, and it would take about a year. [15, p.28] In the mean time, on 17 June, Reaction Motors delivered the second XLR99, and number 2 was transported to the NAA plant to receive it [52, p.82]

By September 1960, both X-15-2 and X-15-3 were at the NAA plant. Number 2 was getting its XLR99 engine installed and number 3 was being rebuilt after its explosion. X-15-1, still equipped with its XLR11 engines,was the only one flying at the Flight Research Center. On 22 September, Navy pilot Lieutenant Commander Forrest Peterson made his first flight in number one, achieving Mach 1.68 and 54,043 feet. He would have gone higher or faster, however both engines shut down prematurely and he was coached to a perfect landing on Rogers Dry Lake - formerly called Muroc Dry Lake. [39, p.122] In November, X-15-2 was back at Edwards with its new XLR99 engine, and on the 15th Scotty Crossfield made the first X-15 flight with the bigger engine. For this flight the traditional NASA instrumentation nose boom was still installed. [26, p.95] He would make two more flights with the XLR99 engine on 22 November and 6 December, completing North American Aviation’s X-15 contractual obligations. This would be Scotty Crossfield’s last research airplane flight. [52, p.82]

Neil Armstrong with hand on the Q-ball airflow direction sensor. Two reaction control system jet orifices can be seen behind the Q-ball housing. NASA photo

On 30 November 1960, a pilot new to the project, Neil Armstrong, made his first familiarization flight in an X-15 using number 1, and on 9 December he made his second flight this time to test the Q-ball air direction sensor for its first flight mounted in X-15-1. [15, p.206] The ball nose and the NASA nose boom were interchangeable, and the nose boom would be used for 27 XLR11 engine flights and the first three XLR99 flights. The ball nose would be used for the last three XLR11 flights and all the rest of the XLR99 flights. [26, p.158] On 7 February 1961, MAJ Robert A. White flying number 1 made the last X-15 flight using the XLR11 engines. The plane was then trucked back to the North American plant for XLR99 installation. The number 1 X-15 had made 21 XLR11 powered flights, and number 2 nine. Seven March marked the first NASA managed flight of X-15 number two with XLR99 engine when MAJ White pushed the craft to Mach 4.43 at 77,450 feet, making it the first manned flight ever to exceed Mach 4. [26, p.102, 110] This was the start of the program to reach the outer limits of the X-15 flight envelope.

In keeping with project goals, on 30 March NASA pilot Joe Walker set a new unofficial world’s altitude record 169,600 feet, wearing for the first time the improved David Clark A/P22S-2 full pressure suit. [52, p.82] Walker experienced two minutes of weightlessness on this flight, the longest period so far. Also, the plane had a 13 cycle per second flutter in the horizontal stabilator that was found to be a combination of the structural resonant frequency of the stabilator, fed by the high gain of the stability augmentation system (SAS). This marked the first time a SAS contributed to a structural resonance, however Walker suspected the cause and was able to stop the resonance by reducing pitch damper gain. [26, pp.110-111] On 23 June, White became the first man to exceed Mach 5, achieving a speed of Mach 5.27 in X-15-2. [39, p.122] Pilots noted that acceleration with XLR99 engine at full thrust was so great that it became difficult to breathe. [15, p.80] The number one X-15 was back at the Flight Research Center with its new XLR99 engine on 10 June 1961, and on 10 August LCDR Peterson made the first flight of the re-engined craft, exceeding Mach 4. [15, p.117] On 9 November, Major White, in the number 2 craft, pushed the envelope to yet another project goal by exceeding Mach 6, the X-15’s design speed. [39, p.123]

Goodbye to the Lower Ventral Fin

By mid 1961 the increasing altitudes of X-15 high altitude flights were requiring increasingly high angles of attack for recovery, and as the angles of attack increased the X-15s became more unstable laterally and directionally. To the point that even with the stability augmentation system at high gain, the planes were marginally controllable. Flight Research Center planners Richard Day and Robert Hoey suspected that the negative dihedral effect caused by the large lower vertical stabilizer was the main contributor to the instability. A number of flight simulator runs convinced them even more, to the point that they convinced station senior managers that a high altitude flight should be made with the lower part of the ventral fin removed to test their theory. On 4 October 1961 USAF pilot Major Robert A. Rushworth made a flight with lower ventral removed - to Mach 4.3 and 78,000 feet. At low angles of attack the handling qualities were about the same as with the full ventral, but a angles above 8 degrees, lateral/directional stability was improved. This success prompted NASA to run a full set of wind tunnel, tests of the X-15 without lower ventral. [26, pp.112-113]

A year later, after completion of wind tunnel testing, Day and Hoey updated the flight simulator with the wind tunnel test results, and a number of simulator flights confirmed the improved handling qualities for high altitude/ high angle of attack reentry flights. On 28 September 1962, FRC pilot Jack McKay made a second flight with lower ventral removed, confirming the wind tunnel tests as well as the theory. FRC then began making most of the high altitude flights without the lower ventral. It was found that with the lower ventral removed, it was possible to make reentries manually without the stability augmentation system; in case of SAS failure. Of the total 199 X-15 flights, 126 flights would be made with the lower ventral removed. [26, p.113] [15, p.176]

By late 1961, the majority of X-15 related aerodynamic, stability and control, loads, and heating tests had been completed, and the three planes were being used increasingly as a high speed, high altitude platform to support testing not related to the X-15 project goals. These included: micrometeorite collection, measuring intercontinental ballistic missile plumes, high-altitude sky brightness measurements, navigational star tracker tests for the space program, horizon definition experiments, air density versus altitude measurements, tests of equipment destined for the space program, tests of a hydrogen fueled hypersonic ramjet engine, and ultraviolet star photography. They were becoming flying laboratories. [17, pp.248-249]

Number Three Comes Back

In late October 1961, X-15-3 was returned to the Flight Research Center hangar equipped with the XLR99 engine and completely rebuilt after its engine test stand explosion. Part of the rebuild included installing the Minneapolis Honeywell Regulator Company MH-96 self adaptive control system. This system had a feature that allowed the pilot to command a pitch, yaw, or roll angular rate about any axis proportional to stick or rudder pedal deflection. It also included a self adaptive damping feature about each axis. This means the system would attempt to maintain a constant ideal damping ratio of oscillations about any axis regardless of flight condition. This it did by sampling airplane dynamic responses to small control surface deflections and then automatically changing the damper gain about any axis to achieve the ideal response. The system not only worked the aerodynamic flight control surfaces but also integrated the out-of-atmosphere ballistic control system with the aerodynamic controls; all under control of either the center stick or the right located sidearm controller. The left located ballistic control handle would not have to be used. Test pilot Neil Armstrong was designated pilot specialist on the MH-96 system, and learned it in great detail. He was also to make the first four flights of X-15-3 to evaluate the adaptive control system. [58, p.100]

X-15-3 was to be used for high altitude investigations, and Armstrong’s four drops were altitude buildup flights, the first of which he made on 20 December 1961 to 81,000 feet and Mach 3.7. His next two flights in January and April 1962 topped out at 133,500 and 180,000 feet respectively. [39, p.123] On his fourth flight in number 3 on 20 April, he reached 207,500 feet at Mach 5.3. During his reentry, Neil was monitoring a feature of the adaptive control system, and did not notice that he had pulled up too abruptly and had actually skipped back up out of the atmosphere where his aerodynamic controls no longer worked. He could only point the airplane in a given attitude with the reaction control system without actually changing its flight path. To make matters worse, he had overshot Edwards and was proceeding over the Los Angeles area, and possibly getting too far from Rogers Dry Lake to make it back. He finally dropped low enough over Pasadena to regain aerodynamic control and made it back to the south edge of the lake bed. On final approach he actually came in below the tops of the Joshua trees at the edge of the lake bed and touched down just 200 feet from the edge. His support vehicles had to drive ten miles to reach him, and he set the record for the longest duration X-15 flight at 12 minutes 29 seconds. It was also the lowest speed landing of the program. [58, pp.103-106] [15, p.216]

On 9 November 1961 Air Force Major Robert M. White became the first man to fly faster than Mach 6, achieving M 6.04 in the number two plane, and exceeding the X-15 speed design goal. In January 1962, Lieutenant Commander Forest Peterson made his last X-15 flight before taking over a Navy fighter squadron. On this occasion, he could not get an engine start in X-15-1 and had to put down on his launch site lake: Mud lake. As a captain, Peterson would later take command of the nuclear powered carrier USS Enterprise, and his last tour would be as a Vice Admiral in charge of the Naval Air System Command. [26, p.115] This writer would have occasion to work with VADM Peterson while carrying out my assignment in charge of developing Navy standard airborne computers during my final tour of navy duty as director of the Naval Embedded Computer Program Office. MAJ White had achieved X-15 design speed in November 1961, and on 30 April 1962 NASA pilot Joe Waker rose to 246,700 feet in the number 1 ship, bettering design altitude. Then on 17 July, White climbed to 314,750 feet in X-15-3, exceeding 50 miles and qualifying for Air Force Astronaut wings. [52, p.82] Three NASA test pilots would eventually exceed 50 miles, but NASA did not award X-15 pilots such wings until many years later. [58, p.110]

Roll-Over at Mud Lake

Flight Research Center pilot Jack McKay was scheduled to make a stability and control flight, with lower ventral fin removed, in X-15-2 on 9 November 1962. His drop site was over Mud Lake, Nevada, and he got a good engine light at 30% thrust. The problem was when he advanced the throttle to 100% the engine did not respond. He could have made the trip back to Edwards at 30% thrust, but there was no telling whether the engine would continue to run, even at 30%. Procedural rules called for him to dump propellants and land on Mud Lake, and he was able to release most, but not all propellants. On his final approach to the lake landing strip, he was in good shape until landing flap actuation. The flaps did not come down, and he had to make a fast touchdown at almost 300 mph rather than the normal 230 mph. The combination of high speed and weight of remaining propellants created a much greater than normal impact load on the rear skids, causing the left skid strut to collapse. If both struts had collapsed the plane probably would have slid straight ahead on its belly, but the plane was now sliding on the left wing tip, the nose wheel, and the right rear main skid. This caused the plane to veer to the left until it was sliding almost sideways, and Jack could see that the right wing tip was about to dig into the lake bed, which would cause the craft to flip on its back. If he was trapped in the sealed cockpit with ammonia vapors seeping in, he knew he would not last long with his remaining supply of breathing oxygen. His reaction was to blow the canopy just before the plane flipped. The problem now was there was no protection between the ground and his head; the canopy was, in effect, his roll-over bar. His head was the first to hit the lake bed, and it hit hard. Fortunately the plane stopped sliding as it flipped. [58, pp.227-228]

X-15-2 after its crash at Mud Lake, Nevada on 9 November 1962. Thanks to its steel and titanium construction, the plane could be rebuilt, but pilot Jack McKay eventually fared worse, passing away prematurely at the age of 52, attributed to injuries from the crash. NASA photo

There was always a rescue crew waiting at a launch lake, and they had to dig into the lake bed under the cockpit for about an hour to get Jack out. Amazingly, he talked to the crew the whole time, and did not appear to be injured, however X-rays at the hospital showed that he had two crushed vertebrae. [58, p.230] Jack was flying again in six months, and made 22 more X-15 flights, but his injuries began to tell on him and pain became increasingly worse. Except for mangled horizontal stabilizers, shredded ventral fin and torn off nose gear, X-15-2 was relatively intact thanks to its steel and titanium structure. The Air Force was developing an experimental supersonic combustion ramjet engine (scramjet) that needed a platform capable of speeds up to Mach 8 to make the scramjet work. The scramjet was an air breathing engine that would run on hydrogen, negating the need to carry a tank of oxidizer, and giving the promise of extending hypersonic flight time. [15, p.305, 313] The Air Force decided to rebuild the number 2 plane, and fit it with external propellant tanks in addition to the internal tanks to give it Mach 8 capability. The experimental scramjet engine would be fastened below the plane in place of the lower ventral fin. [26, p.129]

Simulator flights showed that X-15-3 should be able to climb to over 400,000 feet; the trouble was it would just barely be able to pull out in a reentry from that height. Another problem was it was not easy to precisely top out at a specified altitude. There were too many variables such as accurately maintaining the planned climb angle, precisely controlling engine thrust, and shutting the engine down at just the right second. If things went wrong, the plane could top out 40,000 feet higher or lower than the target altitude, and for this reason 360,00 feet was considered the highest safe target altitude. On 22 August 1963, FRC pilot Joe Walker was scheduled to make a 360,000 foot altitude attempt in the number three craft. It would be the highest man has ever gone in a winged aircraft. He topped out slightly low at 354.200 feet (67 miles), but set a new unofficial world’s altitude record for winged aircraft that would not be exceeded until the advent of the space shuttle 18 years later. While above the atmosphere, the MH-96 adaptive control system could not take automatic command of the reaction control system because of a frozen hydrogen peroxide thruster, and Walker had to manually operate the reaction control system to position the craft for reentry. On reentering he had to pull seven gs, the X-15s maximum design limit. This marked Joe Walker’s last X-15 flight. [58, pp.24-126]

To us, HSFS test pilot Milton O. Thompson was “Uncle Milty”; named after his television namesake Milton Berle. At the lunch table he would regale us with stories of his crop dusting adventures in Mexico as he worked his way through college. For example, he told us of how he hung trapped upside down in his turned over biplane while listening to dripping raw gasoline hissing and sizzling on the hot engine. Thompson would make his first X-15 flight on 29 October 1963 in the number one plane, achieving Mach 4.1 at 74,400 feet. [39, p.123]

X-15A-2: the Phoenix

In May 1963, the Air Force contracted with North American Aviation to rebuild the crashed X-15-2 to be a Mach 8 test bed for the experimental hydrogen fueled, air breathing scramjet engine. It was to undergo enough design changes to be redesignated X-15A-2. [26, p.129] Changes included two large external propellant drop tanks, and a 28-inch extension of the fuselage to hold a liquid hydrogen tank for the scramjet. The external tanks held 1,800 gallons of propellant that would allow an additional 60 seconds of engine burn; enough to get the craft up to Mach 8 where the scramjet could take over. The plan was to consume the external propellant first to get the craft up to Mach 2 at 70,000 feet, where the tanks would be dropped, and the engine would be fed from the internal tanks. [58, p.227, 231] The full external tanks and internal modifications added 35% more weight to the plane, posing an interesting, but doable, challenge to the mother plane pilots. [15, p.266] The bottom of the ventral stabilizer stub provided a convenient place to attach the scramjet, though it would require extending the rear landing skid legs by seven inches to keep the 30-inch diameter engine from scraping on the ground. [26, p.129] The number two plane would be flying faster than it had ever gone before, in fact beyond its design speed, and if not given some sort of heat protection, leading aerodynamic edges would melt or burn up from the frictional heating. The solution was an ablative material developed for intercontinental ballistic missile nose cones. The entire plane was to be coated with thickness varying over the plane in proportion to the expected heating. [58, pp.231-232]

The new configuration of X-15A-2. In addition to a 28-inch fuselage lengthening, changes included: A. cockpit canopy was fitted with oval windows to better withstand heat expansions, B. two external propellant drop tanks, and C. rear skid strut extended by seven inches to clear the scramjet body attached to the ventral fin stub. The left tank was filled with ammonia, and the right tank carried liquid oxygen in one section and helium bottles in another section. NASA drawing

The rebuilt X-15A-2 was accepted by the Air Force on 15 February 1964, and this time government pilots would take over the testing program without North American contract demonstrations. USAF Major Robert A. Rushworth made the first flight on 25 June without the external tanks to test the airworthiness of the basic airplane, and he found the craft handling not much different than its initial configuration, when taking it out to Mach 4.59. [26,, p.131] On his second flight (14 August), also without external tanks, he stepped the speed up to Mach 5.23, but during deceleration at a speed around Mach 4.5 he heard a loud bang. Immediately the nose dropped violently & yawed, and the craft rolled 90 degrees. To make matters worse, the cockpit filled with smoke. He suspected from the sound and location of the bang that the nose gear had extended at Mach 4.5, and the smoke was probably burning rubber. When his chase plane caught up with him, the pilot confirmed that his nose gear was down and the tires looked badly charred. The choice was to either bail out or try to land the plane on tires that would probably disintegrate, and a shock absorbing strut that might not provide any cushion; with probability the plane would break in half. Rushworth and the flight controllers decided to make the landing in which the fuselage held, but there was no rubber to be seen on the bare wheel rims when the craft came to a stop. Investigation revealed that aerodynamic heating had expanded fuselage length so much that the landing gear release cable had inadvertently unlatched the gear release mechanism.[58, pp.233-235]

Flight Research Center test pilot John B. McKay and X-15-3 in which he made his astronaut qualification flight. NASA photo

By late October 1965, various test pilots had thoroughly wrung out the basic X-15A-2 over most of its flight envelope in 11 flights, and the plane was ready for a run with external tanks. On 3 November, Major Rushworth found the plane handled surprisingly well with empty external tanks. The only problem came during the tank jettison test when the ammonia tank parachuted with minor damage, but parachute failure caused destruction of the oxygen tank. [58, p.239] Later, on 6 May 1966, Jack McKay, flying X-15-1, was forced to make an emergency landing on Delmar Dry Lake caused by a broken propellant turbopump casing. Even though the landing was long and McKay ran a few hundred feet off the lake bed into the desert, the plane was undamaged and MCKay took it to the air a month later. Jack McKay made his final X-15 flight on 8 September 1966 in the number 1 craft. Purpose of the flight was to be a test of an experimental horizon scanner to support a non-X-15 project, and evaluation of a modification to the horizontal stabilizer. Premature engine shutdown required abortion of the flight and an emergency landing at Smith Ranch Dry Lake. After that, Jack’s back pains became so severe he could no longer fly, and he retired on disability in October 1971. He had served at NACA and NASA for 21 years, and his premature death at age 52 was attributable to his Mud Lake crash injuries. On 28 September 1965, McKay had flown X-15-3 to an altitude of 55.98 miles, qualifying him as an astronaut. His daughter posthumously accepted the wings in August 2005, 40 years after the flight. [15, pp.155-156, 162-163]

On 1 November 1966, FRC pilot William H. Dana planned to carry seven non-X-15 related experiments to an altitude of 267,000 feet and Mach 5.27 in the number 3 plane, however he overshot his flight plan and topped out at 306,900 feet and Mach 5.46. This would not only qualify Dana as an astronaut, but also marked the last X-15 flight above 300,000 feet. Regardless of the qualification, NASA did not, at the time, give X-15 pilots astronaut wings, they went only to the pilots in the space program. By this time, the three X-15s were serving more as experiment carrying platforms than as research planes in their own right. The next X-15 flight was by Major William J. “Pete” Knight in #2 on 18 November 1966, this time to test handling characteristics of the rebuilt plane carrying full external propellant tanks, and to evaluate a patch of the insulating ablative material on a leading edge. Knight went to Mach 6.33 at 98,900 feet, consuming his external and internal propellants. The external tanks separated well at Mach 2, their parachutes worked, and they were recovered. [58, p.344] From late november 1966 to late April 1967, a new Air Force pilot, Major Michael J. Adams made two flights in # 1 and one flight in # 3 while FRC pilot Bill Dana made one flight in X-15-3. The tests done were about half X-15-specific tests and the other half using the planes as test carrier. Then on 8 May, X-15A-2 was ready for its first flight with dummy scramjet attached. On this flight, MAJ Knight took the plane out to Mach 4.75 at 97,600 feet, then tested scramjet ejection as it was preferred to land with the engine not attached. The explosive bolts caused a clean separation but the dummy engine’s parachute did not work and it plunged into the desert floor; but was repairable. [15, pp305-306]

X-15A-2 about to drop from the mother B-52 with external tanks and dummy scramjet engine attached. The scramjet is the cylindrical shaped object attached to the ventral fin stub. A white protective layer covers the ablative coating material. NASA photo

Aerodynamic heating flights in the number 2 plane were flown at high speeds and low altitude to induce the maximum amount of structural heating. These flights were also a good way to evaluate the aerodynamic effects of the dummy scramjet attached to the ventral fin. FRC pilot Milt Thompson specialized in low altitude flights and heating studies, making 14 flights, more than half of which were heating flights. Leading parts of the plane would get up to over 1,300 degrees, parts would glow cherry red and began to soften. The pilots reported lots of popping and banging due to unequal structural expansions, very much like the sound of a pot bellied stove. [58, p.98] After heating flights, most insignias and marking signs were usually burnt off and had to be repainted. [15, pp.289-290] On 21 August 1967, Major Pete Knight made the first X-15A-2 flight with full ablative coating, and with the dummy scramjet attached, but without external tanks. It was found that the coating caused higher drag, but did not cause much change in handling qualities. Maximum speed was Mach 4.94, and the ablative coating on the ventral fin behind the scramjet had burned away, apparently due to shock waves from the dummy scramjet engine impinging on the fin. [58, p.241]

The next flight of X-15A-2 was to be on 3 October 1967, piloted by Major Knight, and it would be the fastest Flight of the X-15 Program. The external tanks were full, dummy scramjet attached, and the craft covered with heat resistant ablative material. External and internal propellants were totally consumed, and Knight achieved Mach 6.72 at 102,00 feet. This set an unofficial world’s speed record for manned, winged aircraft that would not be exceeded until 1981 when the Space Shuttle Columbia reentered from its first orbital flight. For comparison, Mach 6.7 is about 6,700 feet per second (fps) whereas the speed of a high powered rifle bullet is about 3,000 fps. [26, p.144] During deceleration, Knight made a number of stability and control evaluation pulses to test the dynamics of the airplane/scramjet combination. Heating was so severe on the ventral fin from shock waves from the scramjet that much of the fin’s leading edge was burned away, and the explosive bolts fastening the scramjet got so hot they voluntarily fired, jettisoning the dummy engine. Later calculations revealed that the area had been heated to around 2,700 degrees F. The recovered scramjet body was also severely heat damaged, and if any more of the fin had burned away, hydraulic lines would have been consumed, making the craft uncontrollable. [ 58, p.227, 243-244] This was the closest any X-15 had come to in-flight failure from aerodynamic heating, and FRC engineers admitted that if they had known the heating effects were going to be that bad, they would never have made the flight. In total, the X-15A-2 had made 21 flights, only two of which were with the ablative coating. NASA made the decision to end flight testing of the number two plane, and MAJ Knight’s 3 October flight was its last, it would never fly again. Knight had made 16 flights in all three X-15s, one of which had gone above 50 miles qualifying him for USAF astronaut wings. For 14 years, he was the fastest man alive. [26, p.147]

Loss of Major Adams and X-15-3

On 15 November 1967 USAF pilot Major Michael J. Adams made his seventh flight in an X-15, in this case the number 3 craft in which he had made two previous flights. His mission was to evaluate a new Ames Laboratory Boost Guidance Display and to conduct a number of non X-15 related experiments. His flight plan called for him to climb to 250,000 feet at a maximum speed of Mach 5.1. [26, p.149] On his climb out electrical circuitry in one of the experiments had heavy electrical arching resulting in severe transients in the plane’s electrical supply system that caused the reaction control system, the inertial platform display, and the adaptive control system to malfunction, and causing Adams considerable distraction. It was later found that the malfunctioning experiment had been designed with electrical components not qualified for such high altitude flight. It was also later suspected that components in the MH-96 adaptive control system were damaged by the electrical spikes. [58, p.254]

The X-15’s inertial platform cockpit display indicator had a horizontal and a vertical pointer, in which the horizontal pointer represented angle of attack, and the vertical pointer represented yaw angle. In the case of X-15-3 there was, however, a second display mode for the two pointers - that could be selected by the pilot. The purpose of the second mode was to give the pilot precise pointing error deviations from the planned flight path when orienting the airplane for certain experiments. In this mode, the horizontal pointer represented pitch angle errors and the vertical pointer represented bank angle error. [26, p.196] Adams came quite close to the planned speed and altitude, topping out at 266,000 feet and Mach 5.2 at highest speed. To this point the flight could be called highly successful. Upon nearing experiment altitude, Adams was to switch to the second inertial platform display mode to orient an experiment using an ultraviolet optical instrument to track the plane’s engine exhaust plume. To do this he was to mildly rock the planes wing using the precision roll error display, after which he was to switch the indicator back to the normal angle of attack and yaw angle display. [26, p.149] Later analysis of the instrumentation recordings showed that instead of mild wing rocking, he had made very pronounced bank angles, far more than called for. This would later be a clue that something was going wrong in his instrument interpretations, and it must be realized that it was almost impossible for the pilot to sense airplane attitude by looking at outside references because of the limited cockpit window visibility and the impossibility of seeing reference points on the airplane’s exterior. The pilot was solely dependent on his instruments to know attitude. [58, 38]

Later analysis showed that, as the plane went over the top of its flight path it had rotated 15 degrees to the right of path direction. The ground stations had no way of knowing this because airplane heading did not show on their instrument readouts. All they could see was that the craft was on the planned flight path, but they did not know its orientation. [58, p.256] By half way into reentry X-15-3 was oriented 90 degrees to the path, and by the time it was encountering the atmosphere it was actually flying backward, but this would not be known until later analysis. Upon encountering the denser air the craft went into a Mach 5 spin, and Adams reported three times that he was in a spin. Analysis showed the plane spun from 230,000 feet down to 125,000 feet where Adams managed to recover from the spin and entered a straight dive. There is high probability that Adams could have pulled out from the dive and made a normal landing, but then something else went wrong. [58, p.257]

The MH-96 adaptive control system endeavored to maintain constant damping ratios around the three axes of rotation by monitoring aerodynamic responses to small control surface deflections and varying damper gain to keep those responses the same in all flight conditions. Vacuum tube electronics of the day needed fairly stable power supply and were quite sensitive to power spikes. There is evidence that voltage spikes in the electrical system had destroyed some of the components in the MH-96 system causing it to malfunction. In any event, recordings showed that the system adjusted pitch damper gain to its maximum and that the system actually caused the plane to begin a wild divergent pitching oscillation, and Adams was unable to override the system. The tendency of the adaptive control system to be induced into creating sustained airplane oscillations as a result of abrupt control inputs had also been noted by other pilots, but they had been able to manually reduce system gain. Acceleration forces built up to plus and minus 13 times the force of gravity; at which point the wings and tail surfaces separated, and the craft plunged in pieces into the desert. [14, p.38] Over half of the wreckage could not initially be found, but the main part of the fuselage, with Adam’s body still in the cockpit was located by a chase plane soon after the crash. It could not be determined whether he had tried to use the ejection system. [58, p.38]

Accident investigators were nonplussed when they found the plane had actually reentered tail first, and they began suspecting that Adams had been interpreting the inertial platform readout pointers wrong. One clue was the deep divergence in roll he had made when he was supposed to mildly rock his wings. This could be explained by assuming he had set the instrument to display angle of attack and yaw readings, when he should have had it set to show pitch and bank angle errors. Then it could be assumed he had reset the instrument prior to reentry, but in this case it would have been showing pitch and roll angle errors when it should have been showing angle of attack and yaw angle. This could explain his backward reentry. To confirm this speculation it would be necessary to find the instrument panel recording camera that was mounted on the cockpit canopy and filmed from a position behind the pilot’s shoulder. [58, p.261]

The main part of the X-15-3 fuselage after the 15 November 1957 loss of Major Michael J. Adams. This time the plane was unrepairable, and recovered components were buried at an undisclosed location on the Edwards Air Force Base reservation. NASA photo

A search party eventually found the canopy. It was fairly intact, but unfortunately the cockpit camera had been torn from the canopy and was nowhere to be seen. Now they had to find a small neutral colored object about the size of a VCR cassette tape. The search party of volunteers spread out, line abreast, about ten feet apart and combed the desert floor over hills and valleys for about two miles, until they found the camera. But, another disappointment; the film container was gone. This time they would be searching for something the size of a pack of playing cards. The searchers fanned out again, and unbelievably they found it, fairly intact. Film development revealed that the last few frames had been fogged by light leaks, but the rest of the film showed the flight history of the instrument panel. Investigators confirmed their suspicion that Adams had been interpreting roll error angle as airplane yaw angle, and in so doing had turned the plane completely tail first before reentry. [58, p.262] The final investigation report cited a combination of pilot error and failure of the MH-96 adaptive control system that should have helped Adams recover from the violent longitudinal oscillations, but had actually made them worse. The concept of using a cockpit indicator to represent, by pilot selection, two different quantities also came into question. The Flight research Center did install a temetered heading indicator in the control rooms so that the ground controllers could see deviations from the planned flight heading. [26, p.153] The finding of pilot error must be questioned in light of the fact that Adams had recovered from the spin and had entered a dive from which he could have recovered if it had not been for the failure of the MH-96 system.

Program Termination

Test pilot William H. Dana at far right made some of the F-107 test flights and also made the last flight of the X-15 program. The four to the left are the F-107’s very excellent ground crew that sometimes could get two flights a day out of the F-107. Pilots who flew the F-107 said it was the “best fighter the Air Force never bought,” The author is standing next to Dana. NASA photo

North American Aviation repaired the badly burnt X-15-2 ventral fin, but the plane was never flown again, and it was eventually sent to the Wright Patterson Air Force Base museum. NASA officials could see that the X-15s were being pushed well beyond their design limits, and decided to shut the X-15 program down while they were still ahead. Furthermore, by this time, almost all of the X-15 project objectives had been met and the planes were being used primarily as research platforms for experiments not part of the X-15 project. The X-15 project would only fly another eight missions after the fatal X-15-2 crash, all in X-15-1. [26, p.153] [58, pp.244-245, 265] Flight Research Center pilot Bill Dana made the last flight on 24 October 1968, and again, it was in support of non-X-15 related projects. The main purpose of the flight to carry aloft an experiment to measure the rocket plume signature of a Minuteman II intercontinental ballistic missile, and required that the X-15 drop be coordinated with a Minuteman launch from the Western Missile Test Range at Vandenberg AFB on the California coast. The experiment equipment was mounted in a bay behind the cockpit under a hatch to be opened at highest point of flight and extended above the top of the fuselage. Unfortunately the experiment sensor did not extend due to a burnt out relay in the experiment system. There were also other experiments on board, most of which worked. [15, p.353-357]

In summary, the three X-15s made 199 free drop flights, and were flown by 12 NACA, NASA, USAF and USN pilots. Eight of these pilots flew above 50 miles and qualified for astronaut wings. The Air force pilots: Adams, Knight, Engle, Rushworth, and White were awarded their wings soon after their flights, whereas the NASA pilots: McKay, Dana, and Walker got their wings only in August 2005. In the case of McKay and Walker, it was a posthumous award. Over half the flights exceeded Mach 5, and four of the flights were above Mach 6, the fastest flight being Mach 6.7 by Major Pete Night on 3 October 1967. The highest altitude achieved was 354,200 feet by Joe Walker in X-15-3 on 22 August 1963. It was calculated that the X-15 had the sufficient fuel to achieve altitudes near 450,000 feet, but it was also calculated that reentry heating would probably destroy the aircraft. X-15-1 was the first and last to fly in the program, and the only one that all 12 pilots flew. It made 142 takeoffs under the B-52s, and 81 free flights. X-15-2 went 97 times aloft under B-52s , and made 53 free flights, 31 before rebuild into the X-15A-2, and 22 after. X-15-3 was carried aloft 97 times, and made 65 free flights. [58, p.273] Amazingly each X-15 had logged only about ten hours of free flight time by the end of the program. Regardless, the planes were beginning to wear out, primarily due to the amount of disassembly and ground testing needed to prepare them for the next flight. [58, p.223]

Today the number 1 plane is at the Smithsonian Air and Space Museum, and number 2 is at the US Air Force Museum near Dayton, Ohio. The wreckage of number three was buried somewhere on Edwards Air Force Base, and a replica stands upon a pylon near the front door of the Flight Research Center.

End Results

The X-15 project had a number of significant findings and contributions to aeronautical knowledge. Perhaps the most important finding was that manned, human piloted flight could be done at hypersonic speeds, and airplane structures could be made to withstand the rigors of such flight. Other findings of note included:

• Flight results were in fairly good agreement with wind tunnel predictions regarding hypersonic airplane coefficients and stability derivatives. It showed that existing hypersonic wind tunnels could fairly accurately predict hypersonic airplane characteristics. The reverse side of this was, in case of non-agreements, X-15 findings were used to improved wind tunnel technology and techniques. [58, p.208]

• There was generally good agreement between predicted and actual measured airplane structural stresses thus verifying analytical techniques. However, aerodynamic heating flight measurements were slightly lower than predicted. [58,p. 210]

• Aerodynamic heating, especially when one or more shock waves impinged at a particular spot on the skin for sustained period (few minutes) could cause the skin to burn through. Also small skin gaps, such as around access doors, could admit jets of superheated air that burned wiring and aluminum fittings. [58, p.214]

• Artificial stability augmentation was found to be almost essential for high speed flight, even with the most experienced X-15 pilots. Handling qualities became very poor without stability augmentation, especially at higher angles of attack.[58, p.95-96]

• The lower ventral fin was found to help guard against roll coupling, but adverse dihedral effects of the lower ventral contributed to rolling instability especially at high angles of attack. The second half of the flight program was done without the lower ventral fin. It was most effective in contributing to static lateral/directional stability at subsonic speeds. [58, p.208]

• The MH-96 adaptive control system in X-5-3 improved out-of-the-atmosphere handling qualities, and made transition from aerodynamic controls and reaction controls easier. However, the MH-96 system could be induced into creating sustained airplane oscillations by abrupt pilot control inputs. [58, p.209] It was found that MH-96 adaptive damping system failed on 25% of X-15-3's free flights. [17, p.321] Most of the failures apparently being due to vacuum tube electronics susceptibility to voltage spikes.

• The ground-based analog flight simulator accurately depicted X-15 handling qualities and could easily be refined and updated from X-15 flight testing results. [58, p.209]

• The original X-15 inertial reference system was unreliable, and pilots had to depend heavily on ground tracking stations for, track, flight profile, speed, and altitude information relayed up to the pilot. [58, p.212]

• The stability augmentation damping system in planes 1 and 2 was marginally reliable, but was essential to controllable flight at altitudes above about 200,000 feet. A partial solution was to add a redundant system that was tested on flight 50. [58, p.213] Again, a part of the problem can be traced to analog electronics susceptibility to the rigors of hypersonic flight.

• There were eight XLR99 rocket engine failures during the flight testing program. The engines were usually reliable at 100% thrust, but often quit when throttled back to 30% thrust. Project engineers eventually set 40% as the minimum useable thrust. [58, p.222]

The X-15 program produced more than 750 reports and research papers. [58, p.266] Even those that were classified found their way to private industry as government furnished contract information. Other X-15 project contributions to aeronautical and space knowledge included:

• Most of the techniques, processes, and procedures established for High Range operation were passed on to the Project Mercury. Gemni, Apollo, and Space Shuttle programs. In particular the methods of tracking and communicating with spacecraft and telemetering flight and biomedical data to ground sites. Much of this knowledge was transferred by the move of key HSFS personnel to the space program. For example Walter Williams the first director of the High Speed Flight Station went on to direct launch and flight operations for Project Mercury. This included establishing and operating the Worldwide Tracking Network and recovery operations for manned space flight missions. Gerald Truszynski, who started the Instrumentation Division at HSFS, and later developed the X-15 tracking range, moved to NASA Headquarters where he developed the NASA Deep Space Tracking Network. [58, p.58]

• The X-15 project produced the world’s first really practical and useable pilot’s full pressure suits. The David Clark A/P22S-2 suit evolved into the standard U.S. high altitude pressure suit, used in projects Mercury, Gemni, Apollo and space shuttle use.[26, p.207]

• The X-15 project was the first aircraft use Inconel X alloy, a very difficult metal to use, and fabrication techniques developed by North American Aviation were made available to the aircraft industry in general.

• The X-15 auxiliary power units (APU) were the among first aviation auxiliary power units, and General Electric performed ground breaking research and development in devising APUs that eventually worked very well in the harsh X-15 flight environment.

• The space shuttles were a direct user of much of the X-15 generated knowledge. [58, p.270]

In addition to leading the way in hypersonic aerodynamics knowledge and aerodynamic heating structural protection, The approach and landing techniques developed by the X-15 project were followed in the space shuttle program. For example, initial planning for the space shuttle called for a special landing engine, however X-15 experience in making unpowered landings convinced the space shuttle managers that a landing engine would not be required and that the orbiter could be landed as a glider. [26, p.93]

• A number of the experiment packages carried aloft in the X-15 were to gather data in support of the Apollo moon project. In many cases they were prototype Apollo systems, subsystems, or components. [58, p.270] A specific example is the testing of the Apollo star tracker. [26, p.167]

Neil Armstrong, center, was the leader of the High Speed Flight Station’s “Hi Hos” singing group. This writer is second from left, singing tenor. From HSFS X-PRESS newsletter Dec 1956

All was not work and “nose to the grindstone” at the High Speed Flight Station. One of our diversions was a small singing group led by Neil Armstrong, who not only had a very good singing voice but was also quite musically talented. We called ourselves the “Hi Hos”, the “Hi” coming from “High” in High Speed Flight Station. We would perform at parties, hospitals, and nursing homes, especially during the holiday season.

Neil Armstrong, center, was the leader of the High Speed Flight Station’s “Hi Hos” singing group. This writer is second from left, singing tenor. From HSFS X-PRESS newsletter Dec 1956

The Neil A. Armstrong Flight Research Center in 2016. The original High Speed Flight Station is at center, and the buildings below it are extensions built over the years. Cover of the March 2014 Neil A. Armstrong FRC X-PRESS News Letter

On 26 March 1976, the Flight research Center was renamed The Hugh L. Dryden Flight Research Center in honor of Dryden who served as NACA Director from 1947 to October 1968. and then as NASA Deputy Director from that time until his death in 1965. Then on 1 March 2014, the Hugh L. Dryden Flight Research Center was renamed The Neil A. Armstrong Flight Research Center.

Since the last X-15 was retired in 1968, no other manned hypersonic aircraft, other than the space shuttles, have been built anywhere in the world in the intervening 48 years. Now in 2016, the idea of manned hypersonic flight is being revived. Christian Davenport, in a 28 March 2016 article in The Washington Post told how “Lockheed Martin joins race to build hypersonic aircraft.” The need for such planes, he noted, were being dictated by emerging weapon systems and surveillance requirements. NASA was mentioned in the article, but there was no mention of the X-15 project. [9] It was as if manned flight faster than Mach 5 was being invented anew. Let us hope, at some point, the hundreds of reports on the X-15 project are dusted off, opened, and reread. The project, done almost 50, years ago was declared at the time the most useful, and highest payoff of all the experimental research airplanes, and perhaps it will continue to pay off. We still owe a lot to “those magnificent men in their flying machines,” and the golden age of experimental research airplanes.