THE NF-104A AEROSPACE TRAINER

With the advent of manned spaceflight in the early 1960s, the US Air Force foresaw an important role for its pilots in space, not only as part of the X-15 program but also as astronauts on Gemini and Apollo missions, manning an orbital USAF outpost, and flying military shuttle-like vehicles starting with the X-20 Dyna-Soar (which will be described later). Thus in 1962 the USAF Experimental Flight Test Pilot School at Edwards Air Force Base became the Aerospace Research Pilots School (ARPS) and its training for armed forces test pilots was expanded beyond the traditional aviation curriculum to include an 8-month aerospace course involving spacecraft operation. In line with the school’s change in scope, it soon required a high-performance but low-cost training aircraft that would be able to fly far up into the stratosphere. With such a plane the flight profiles of the X-15 and X-20 could be rehearsed in order to enable test pilot students to familiarize themselves with flight in very rarefied air, where the standard aerodynamic control surfaces become ineffective and control thrusters are needed. Re-entry is a particularly dangerous phase that leaves little margin for pilot error. For the X-15 and indeed any spaceplane, the correct orientation for re-entry into the atmosphere is of paramount importance because only the properly shielded part of the aircraft is able to protect it from the extreme temperatures that it will experience. In the early 1960s there were no computers that could take care of such delicate maneuvering and were also small enough to be accommodated inside an aircraft. The X-15 was flown manually all the way, and it would be the same for the X-20 once this had been inserted in orbit by its launch rocket. The ARPS training planes would also familiarize students with zero – gravity, since during the unpowered ascent and descent in a near-vacuum they would effectively be weightless.

North American proposed a modified version of their X-15 design, adding another cockpit for an instructor and a conventional undercarriage with wheels instead of the usual skids to enable the trainer to take off under its own power. But the X-15 was a complex, expensive machine. A more cost-effective solution was found in the shape of an F-104 Starfighter jet modified for mixed propulsion. An additional benefit of a modified F-104 was that it would enable students to rehearse reigniting a jet engine after a high-altitude parabohc flight during which it would either be deliberately shut down or left to flameout by being starved of oxygen.

The ARPS had already been using standard production Starfighters to simulate the very steep, low-lift and high-drag glide approaches of the X-15 and the planned X-20 Dyna-Soar. This involved climbing to an altitude of 3.7 km (12,000 feet), throttling the jet engine back to 80%, ‘dirtying up’ by extending the flaps, speed brakes and undercarriage to maximize the aerodynamic drag, and then establishing a 30 degree dive. This resulted in a very steep, low lift-over-drag descent similar to that of a spaceplane. (The lift-over-drag, or L/D, ratio is a measure of a plane’s aerodynamic efficiency: a ‘dirtied-up’ Starfighter had a ratio of 2.2, meaning the lift force its wings provided was only 2.2 times the amount of drag the aircraft caused. In contrast, a ‘clean’ Starfighter had an L/D of 9.2. The L/D of a typical airliner is 17, and that of the X-15 was about 4 during its glide phase.) The pilot would pull the nose up for the landing flare a mere 460 meters (1,500 feet) above the ground, which left very little room for error. It was a risky profile but it did prepare pilots for landing future spaceplanes.

The school’s new F-104 space trainers would need to be equipped with a reaction control system similar to that tested by the X-1B and used operationally by the X-15, to control the aircraft at high altitudes and allow students to rehearse attitude control for spacecraft. NASA already had experience in this, having modified an F – 104A in 1959 to use a hydrogen peroxide reaction control system. Following zoom climbs to altitudes up to 25 km (83,000 feet) this gave the NASA Starfighter controllability in the rarefied upper atmosphere. To achieve higher altitudes the ARPS trainer would have an auxiliary rocket engine in the tail.

In 1962 the ARPS awarded Lockheed (the Starfighter manufacturer) a contract to modify three F-104A single-seat fighters for the dedicated role of aerospace trainer. Three existing aircraft were subsequently taken out of long-term storage, relieved of all unnecessary equipment (such as the cannon) to reduce their weight, then equipped with improved instrumentation for high-altitude flight. The reaction control system was based on that of the X-15 and consisted of four 500 Newton thrusters for pitch, four 500 Newton thrusters for yaw and four 190 Newton thrusters for roll. As on the X-15, two thrusters would fire in parallel for every impulse in each direction. The thrusters ran on hydrogen peroxide from a dedicated tank and the pilot controlled them using a handle mounted on the instrument panel. The wings were extended at their tips both to make room for the roll thrusters and to increase lift in order to compensate for the modified aircraft’s greater weight. The standard vertical fin and rudder were substituted by the larger versions of the two – seat F-104 to increase their effectiveness in the thin air of the high stratosphere, and the fiberglass nose radome and its radar were replaced by an aluminum cone that housed the pitch and yaw thrusters. A battery was added to run the onboard systems when the jet engine cut out. A long nose probe was installed to measure the angle of attack and sideslip of the plane with respect to the airflow without disturbance from the flow around the body.

A normal jet engine only operates with subsonic air flowing into it. On the basic Starfighter, shock cones were installed in front of the air intakes to guarantee that at supersonic speeds a shock wave formed. This slowed the air passing through it into the intakes to subsonic speed. The inlet shock cones on a standard Starfighter created a proper shock shape up to Mach 2 but extensions were fitted to these cones because the new trainer would fly faster than that. Because the jet engine would not operate in very thin air and thus could not provide ‘bleed’ air to pressurize the cockpit at high altitude, an additional pressurization system was needed. The pilot would be wearing a pressure suit that would inflate in an emergency but an inflated suit would seriously hamper the precise control required from the pilot during normal operation, so it was decided to fully pressurize the entire cockpit using nitrogen from an added gas tank (oxygen would have offered the benefit that the pilot could breathe it, but would have been a serious fire hazard).

The F-104As were equipped with a standard J79 jet engine that gave the plane a normal maximum thrust of 43,000 Newton at sea level, which could be increased to

67,0 Newton on afterburner. For the required extra boost a compact Rocketdyne AR2-3 (LR121-NA-1) rocket engine was installed at the base of the vertical tail, just above the jet’s exhaust. It was canted slightly so that its thrust was aimed through the plane’s center of gravity and thus would not continuously push the nose down. This engine ran on a mixture of standard JP-4 kerosene jet fuel drawn from the aircraft’s standard fuel tank and 90% concentrated hydrogen peroxide oxidizer. It provided a thrust of 27,000 Newton and could be restarted and throttled in the 50% to 100% range using a specific throttle lever on the left side of the cockpit. The aircraft carried sufficient oxidizer for about 90 seconds of full-thrust rocket operation. With the need

An NF-104A. Note the long nose probe, wing extensions and the large rocket engine in the tail [William Zuk].

to replace some of the AR2-3’s parts after an hour of operation some 40 flights could be made before a minor overhaul was required (a major overhaul was required after 2 hours and the total life of an engine was 4 hours). The heavily modified Starfighter design was designated the NF-104A (with the ‘N’ standing for ‘Nonstandard’) AeroSpace Trainer (AST).

Every aircraft has a so-called service ceiling, which is the maximum altitude at which it can operate under normal conditions and whilst in steady, horizontal flight. To reach altitudes that are well beyond their service ceiling for brief periods of time jet fighters use so-called zoom climbs in which the plane accelerates horizontally to great speed and then pitches up into a steep climb. Zoom climbs enable aircraft to exploit the good performance of their jet engine at relatively low altitudes to build up speed which can then be traded for height on the way up (the thrust rapidly diminishes due to a lack of air as the altitude increases). The NF-104A pilots used a similar approach but they had the benefit that their additional rocket engine would not lose thrust in the thin air of the stratosphere. An experienced NF-104A test pilot would typically use his jet engine with afterburner to accelerate to Mach 1.9 at an altitude of 11 km (35,000 feet), then ignite the rocket engine to full thrust. Shortly afterwards he would reach Mach 2.2, whereupon he would pitch the aircraft sharply up into a 70 degree climb at 3.5 G. As the amount of cooling air flowing through the engine dropped with the decreasing atmospheric density the jet engine’s temperature would gradually increase. To prevent this from reaching levels that would impair the engine’s structure the pilot would start to throttle down the afterburner at an altitude of about 21 km (70,000 feet) and completely shut down the engine at around 26 km (85,000 feet). After the rocket engine had depleted its hydrogen peroxide the aircraft would continue to climb ballistically farther into the stratosphere until its vertical

rate reached zero. Then it would fall back into the denser air, where the jet engine would be restarted.

During the ballistic climb and descent the aircraft and pilot would effectively be weightless. The pilot remained strapped in but would feel an absence of body weight against his seat, and loose objects would fly through the cabin. However, he had little time to enjoy this experience or the view because during this phase he would need to use the reaction control system to nose the aircraft over through 140 degrees in order to cancel the climb attitude and push the nose 70 degrees down for re-entry into the atmosphere whilst maintaining zero roll and yaw. This ensured a stable position in the thicker atmosphere and also that sufficient air would flow into the jet’s intakes to windmill its compressor and thereby enable it to be restarted. A poor entry attitude would result in a pilot soon finding himself in an out-of-control aircraft and unable to restart the engine (as occurred to Chuck Yeager during a particularly hairy NF – 104A flight, as we will see later).

The Bell X-2 had reached altitudes very similar to those planned for the NF-104A but unlike the long, thin Starfighter with its tiny wings the X-2 was an excellent and stable glider designed to land without engine power. An additional complication that did not affect the all-rocket X-2 was the gyroscopic effect of the still rapidly turning turbine and compressor of the extinguished jet engine: this effect resisted the reaction control thrusters, and required careful compensation by the pilot whilst executing the pitch over maneuver.

The F-104 itself was, even by today’s standards, an impressive high-performance aircraft and with the additional rocket engine it truly became a near-spaceplane. But it was definitely something for experienced pilots (as all ARPS students were). The normal Starfighter was so difficult to handle that it had gained the nickname ‘Widow Maker’. The West German Air Force bought 917 of them in the early 1960s and by 1976 had lost 178 in accidents. The added rocket engine and the need to operate the reaction control system in the high stratosphere made it even more dangerous to fly. During the steep climb the plane’s attitude and the pilot’s rigid helmet meant that the horizon could not be seen, so all critical stability control and maneuvering had to be performed on cockpit instrumentation only. Furthermore the NF-104A was a single­seat aircraft so there was no instructor in the back ready to take over if things went wrong. To ready a student for this challenging aircraft he would first make a flight in a standard F-104 with his pressure suit fully inflated to familiarize himself with the movement restrictions this would impose in an accidental cockpit depressurization. Next he would make a zoom flight in a conventional Starfighter trainer supervised by an instructor in the rear seat. Following 4 hours of rehearsal in a flight simulator he would execute three solo zoom flights in a standard F-104 while being coached by an instructor in the back of an accompanying two-seat trainer. During these flights he would gradually build up the climb angle: 30 degrees on the first flight, 40 degrees on the second, and finally 45 degrees. Only then would he be deemed ready to zoom the rocket-equipped NF-104A, and even then the maximum allowed climb angle was 50 degrees (the optimum for reaching extreme altitudes was 70 degrees).

The first NF-104A (56-0756) was tested by Lockheed’s test pilot Jack Woodman and Major Robert W. Smith of the Test Division of the Air Force Flight Test Center at Lockheed’s factory near Palmdale Airport, prior to its formal handover to the Air Force. During this phase Smith broke the standing altitude record by zooming up to an astonishing altitude of 36,230 meters (118,860 feet) on 22 October 1963. During this flight he managed to keep the aircraft under control even though all three axes of the reaction control system had accidentally been wired incorrectly! After acceptance by the Air Force, and during the next phase of testing at Edwards, Smith surpassed his record by achieving an altitude of 36,800 meters (120,800 feet) on 6 December, and the unpowered parabolic arc provided no less than 73 seconds of weightlessness. Nearly half a century later, Smith’s achievement still stands as the highest altitude ever achieved by a US aircraft taking off from a runway. Although the plane left the ground under its own power (unlike the X-planes) and the altitudes achieved were accurately recorded by ground stations equipped with radar and powerful telescopes, both records remained unofficial because the Air Force had not requested the flights to be monitored by the International Aeronautical Federation.

The second NF-104A (56-0760) was delivered 25 days after the first, and the third and final aircraft (56-0762) on 1 November 1963. Unfortunately the third plane was lost barely a month later, on 10 December, when it crashed during a flight piloted by Chuck Yeager, who was Commander of the Aerospace Research Pilots School at the time. According to him, he was unable to push the nose back down once he reached the zenith of the zoom climb, possibly due to a malfunction of the reaction control system, causing the Starfighter to go into a disastrous flat spin at an altitude of 33 km (109,000 feet). During a normal descent the aircraft would be orientated (using the reaction control system) so that air would flow into the jet engine’s intakes and make its compressor windmill, thus providing hydraulic pressure for activating the control surfaces and enabling the engine to be restarted. However, in the flat spin no air was flowing into the engine, so there was no power to control the ailerons, elevators and rudder and no means to restart the engine. As the aircraft plummeted from the sky, Yeager had no option but to abandon the aircraft. He ejected just 2.6 km (8,500 feet) short of hitting the ground, was struck by his own discarded ejection seat on the way down and was badly burned by its glowing solid rocket motor, but managed to land by parachute. This is depicted in the movie The Right Stuff, although it shows him flying a standard F-104G without a rocket engine.

The ensuing Air Force investigation cleared Yeager of responsibility for the crash, blaming the accident on an aircraft malfunction. However, the NF-104’s primary Air Force test pilot Major Robert Smith, who had trained Yeager to fly the profile, insists that Yeager simply did not perform a proper zoom climb, pulling up to the full 70 degree climb angle too slow and too late. This meant he ran out of speed before reaching the intended maximum altitude, was too late trying to nose the aircraft down, and hence began to fall with the nose still 70 degrees up. Yeager may have believed he was still climbing because the rocket was still operating, but by then the thrust of the rocket alone was insufficient to prevent the plane from falling back essentially tail first. Still in relatively dense air, no amount of reaction control thrust could have maneuvered him out of the attitude in which it was impossible to restart the jet engine. According to Smith, Yeager’s fame and influence meant that the investigation ruled in his favor and unjustly labeled the NF-104A a dangerous aircraft. The differing accounts of this incident, Yeager in his famous autobiography and Smith on his highly informative NF-104A website, don’t even agree on the purpose of the disastrous flight: according to Yeager it was part of his investigation of a known pitch-up problem of the aircraft (caused by the T-tail being masked by the wings at high angles of attack, preventing the airflow from reaching the horizontal stabilizers; the same problem that troubled the British SR. 53) while Smith insists Yeager was merely trying to break the altitude record and had been assigned by the Air Force to fly the NF-104A solely for this purpose.

After the accident a restricted flight regime was enforced to ensure safe flights for the ARPS students using the two remaining NF-104As, part of which was imposing a limit of 50 degrees on the chmb angle (again, unjustly according to Smith, which in his opinion left the students with little opportunity to experience the peculiarities of reaction control at really extreme altitudes).

The remaining aircraft were used to train students, but not very often and only for zooming flights to relatively low altitudes. The dangerous hydrogen peroxide caused some trouble, as was to be expected from experiences with earlier hydrogen peroxide powered rocket aircraft. Once a tail tank ruptured on the ground and another time a small explosion occurred in a wing while in flight, both caused by hydrogen peroxide reacting with metal aircraft parts. Modifications where made to prevent recurrences but the 56-0756 suffered an inflight rocket motor explosion in June 1971 owing to a hydrogen peroxide leak. The rocket engine and most of the rudder where blown off but the student was able to land safely. The seriously damaged aircraft was scrapped.

By then, however, the Air Force’s human spaceflight ambitions had withered: the X-15 program was over, the X-20 had long since been canceled, and the task of the planned Manned Orbiting Laboratory (MOL), namely military reconnaissance, could be done better and at far lower cost by unmanned spy satellites; MOL was canceled in 1969. The Space Shuttle would not fly for another decade and it would be run by NASA. The remaining NF-104A was therefore retired, mounted on a pole and placed outside the Air Force Test Pilot School where it can still be seen today.

Various parts of this aircraft, including the extended wing tips and the metal nose cone, were loaned to Daryl Greenamyer for his civilian aviation record attempts with a highly modified Starfighter that was based on equipment from various F-104s. In 1977, after a practice zoom flight working up to his altitude record attempt, one main wheel of the plane’s undercarriage did not completely deploy and Greenamyer had to eject. The NF-104A parts were lost along with the rest of his aircraft.

Several test pilots who flew the NF-104A before the planes were handed over to the ARPS experienced severe difficulties in controlling the aircraft. To achieve high altitudes and to allow sufficient time in near-vacuum for the very large (140 degree) change in pitch angle required the pilot to fly a very precise zoom maneuver. The initial speed, the fast pull-up maneuver to the 70 degree climb angle, and maintaining this angle, were all extremely important. Near the peak of its parabohc trajectory the NF-104A moved from the aerodynamic control region into the space control region and back in less than a minute, giving the pilot little time to transition from the well – known aerodynamic controls to the less familiar reaction controls and back again.

Nevertheless, according to the Air Force’s primary test pilot, Robert Smith, it was

not a particularly dangerous aircraft to fly as long as the pilot flew the proper zoom trajectory, had sufficient understanding of the peculiarities of reaction control at such altitudes, and did not attempt to push the aircraft beyond its established boundaries. Indeed, X-15 test pilot Bob Rushworth flew the NF-104A to the impressive altitude of 34 km (112,000 feet) without trouble on his first and only flight in it. In total some 50 pilots flew the NF-104A during 302 flights and accumulated a total of 8.6 hours of rocket engine operation.