MIXED UP; COMBINED JET/ROCKET INTERCEPTORS

Of course, as some German and Russian projects had explored during the war, there was a way to combine the fantastic rate of climb and high-altitude performance of a rocket and the endurance and range of a jet in a single aircraft.

After the failed attempts to create mixed-power ramjet/rocket aircraft during the war, the Russians did not resume their interest until the mid-1950s. The MiG 19SU (also known as the SM-50) was a MiG 19S equipped with a U-19 rocket boost unit based on a 29,000 Newton S3-20M engine running on nitric acid and kerosene. This was specifically intended for intercepting the elusive Lockheed U-2 reconnaissance aircraft. This CIA spy plane with its extremely long wings could fly at an altitude of 21 km (70,000 feet), beyond the reach of normal fighter aircraft. The rocket booster was a self-contained unit that required only electrical and mechanical attachment to the aircraft. It could be carried on the belly of a MiG 19S which, apart from some structural reinforcements and the interfaces, was otherwise a standard version of this jet fighter. Ground clearance was narrow with the booster under the fuselage, so the aircraft needed a smooth runway and the rotation immediately prior to taking off had to be shallow. From November 1957 to February 1958 five prototypes made a total of 44 rocket propelled flights that demonstrated the aircraft could reach a maximum altitude of 24 km (79,000 feet); an impressive improvement of the normal ceiling at 17.5 km (57,400 feet). This altitude was reached using a ‘standard intercept’ profile in which the MiG used its jet engine to get to 14 km (46,000 feet) and then ignited its rocket engine to make a quick jump further into the thin air. However, the total time to reach its peak altitude was 12 minutes, which was not particularly spectacular, and it could not remain there for any length of time. The wings and control surfaces were never designed to operate in rarefied air, and above 20 km (66,000 feet) it was about as maneuverable as a brick and difficult to point in the right direction for firing its guns and missiles (which in those days had to be aimed much more accurately than their modern, highly autonomous descendants). In addition, the aircraft was prone to stalling at high altitude and the lack of oxygen tended to make its turbojet flame out. This meant that even under the control of an exceptionally skilled pilot, the aircraft could only make a running jump up at a U-2 in the hope of firing a few shots in the general direction of the target at the top of its parabolic trajectory, before it tumbled down again. In short, the chance of a successful intercept was remote. A number of MiG 19P fighters which had the ‘Gorizont-1’ (Horizon) radio datalink for automatic guidance to a target by ground control were also modified for the U-19 booster but these MiG 19PUs were not very effective either. By then surface-to-air missiles were capable of shooting down a U-2; as was shown on 1 May 1960 with the downing of Gary Powers by an S-75 (SA-2) missile. (A second missile struck a high-flying MiG which was also attempting to intercept this particular U-2.) After this missile success, the MiG 19SU and MiG 19PU aircraft were retired.

In 1954 MiG also initiated the development of a mixed-power, short-range point defense interceptor variant of the Ye-2 swept-wing jet prototype fighter they were in the process of creating (the delta-winged version was the MiG 21, a famous aircraft type that became a huge success after its introduction in 1959). The new aircraft was powered by a Tumansky AM-9Ye turbojet that delivered a thrust of 37,000 Newton with afterburner and a Dushkin S-155 rocket motor whose maximum thrust was also

37,0 Newton. The S-155 ran on (non-hypergolic) nitric acid and kerosene, which meant it could draw fuel from the same kerosene supply as the turbojet engine. This plane was named the Ye-50 and three prototypes were built. The standard Ye-2 was already capable of flying at Mach 2 but the rocket engine was to increase its rate of climb as well as its maximum altitude and maximum speed. The S-155 engine and the tanks for the hydrogen peroxide running the propellant pumps were fitted in the tail, with the rocket nozzle located just above the jet engine exhaust.

The first prototype, the Ye-50/1, flew on jet power on 9 January 1956 under the control of test pilot Y. G. Mukhin, and first used its rocket engine on 8 June that same year. It made a total of 18 successful flights, three of which used combined jet and rocket power. On 14 July the Ye-50/1 made a routine checkout flight after its jet had been replaced but the new engine failed and the plane crashed. The pilot was unhurt but the plane was badly damaged and detailed inspection showed that it was beyond repair. Test flights could not resume until a second prototype of a slightly improved configuration was built. This Ye-50/2 aircraft was under construction from 18 July to 7 December 1956, and on 3 January 1957 made its maiden flight with test pilot V. P. Vasin. No fewer than 25 flights were carried out during the first half of the year, and the plane attained a maximum altitude of 25.6 km (84,000 feet) on 17 June with the help of rocket thrust; considerably higher than the normal ceiling of the standard Ye-2/MiG 21 of about 19 km (62,000 feet). On that same flight the Ye – 50/2 also reached a maximum speed of 2,460 km per hour (1,530 miles per hour), or Mach 2.33, which was about one-third of a Mach number faster than when flown under jet power alone. The Ye-50/2 was joined by the Ye-50/3, which featured a longer forward fuselage in order to increase the internal fuel load. This made its first flight in April 1957. After nine successful flights disaster struck on 8 August when the rocket engine exploded and the plane fell out of the sky. Unfortunately the ejection seat failed and test pilot, N. A. Korovin, was killed when the plane hit the ground. This curtailed the Ye-50 test flights.

In the meantime MiG’s Gorky factory had been ordered to build a batch of 20 pre-production aircraft in preparation for operational deployment as the MiG 23U (not to be confused with the trainer version of the MiG 23 ‘Flogger’ jet which had the same designation but was a totally different aircraft developed a decade later). None were ever built, however, because the Dushkin OKB which built the S-155 rocket motors had in the meanwhile closed down. In any case, the Air Force had seemingly lost its enthusiasm for mixed-power interceptors, at least in part because of the difficulties associated with operating and maintaining the rocket engine. Furthermore, the sheer vastness of the Soviet Union meant that a great number of aircraft would have been needed even to defend a limited number of high-value targets.

In parallel with the MiG Ye-50, a mixed-power version of the Yakovlev Yak-27 twin-engine fighter was under development. This Yak-27V first flew in May 1957 and operated the standard two RD-9Ye turbojets, each with an afterburning thrust of 38,000 Newton, in combination with an S-155 engine mounted in the tail. With this rocket boost the prototype managed to zoom to 25 km (16,000 feet) during its 2- year test program before closure of the Dushkin OKB and diminishing support from the Air Force curtailed further development.

It appears that in addition to Dushkin, the Kosberg rocket design bureau had been working on rocket engines based on oxygen and kerosene propellants for both the MiG Ye-50 and the Yak-27V but it is not clear whether these were ever flown.

These two planes proved to be the Soviet Union’s final attempts at a mixed-power fighter interceptor, because the concept was soon rendered obsolete by improve­ments in jet engine technology.

The US also experimented with mixed propulsion to satisfy a requirement issued in 1945 for a short-range interceptor. Operating in the immediate vicinity of its base this new aircraft was to provide the last line of defense against incoming high-speed bombers, and the specifications emphasized high speed, high flight ceiling and rapid climb. Bell proposed a weaponized, delta-wing version of its X-l supersonic research rocket plane. Republic offered its AP-31 mixed-power aircraft with inversely tapered wings (wider at the tip than at the root in order to avoid wingtip stalls at low speeds) and the capability to vary the angle of attack of the wing with respect to the fuselage in flight (for a high angle of attack and thus high lift during take-off and landing, and a low angle of attack and low drag for high-speed flight). Northrop proposed the XP-79Z, which would be similar to the aforementioned jet-powered XP-79B except that it would be based on a rocket motor (something other than the complex Rotojet that had been planned for the original XP-79). Convair offered a swept-wing aircraft with a very unusual ‘ducted rocket’ powerplant. This was basically a ramjet with several small rocket motors inside the combustion chamber. As well as delivering their own thrust, the rockets would act as igniters for the ramjet. Another four rocket engines, each with a thrust of 5,300 Newton, would be fitted outside the rear fuselage to get the plane into the air and to sufficient speed to enable the ramjet to function. A small Westinghouse J-30 jet engine would provide electrical and hydraulic power, and fly the aircraft back to base once it ran out of rocket and ramjet propellant (it was later proposed that this jet be replaced by a midget racing car engine for supplying power only). As it was to take off using a jettisonable dolly, the plane would be fitted with a lightweight undercarriage for landing only.

In April 1946 Convair was announced to have won the contest by Headquarters Air Material Command of the Air Force and the interceptor project was designated XP-92. Wind tunnel tests soon showed that instead of swept wings the aircraft would behave better with delta wings similar to those developed by Alexander Lippisch for the Me 163 (after Germany’s defeat Lippisch had been relocated to the US as part of Operation Paperclip). The pilot would operate the controls in the prone position in a cockpit that was mounted on the nose of the aircraft like a spike projecting from the circular ramjet air intake, and which could be completely ejected in an emergency. The four external rockets would now each have a thrust of 27,000 Newton and there would be a total of sixteen 220 Newton rocket engines inside the ramjet.

The design emerging was not particularly appealing to the eye: a tubular vehicle with a conical nose and a large triangular fin and delta-wings that made it look like a small dart. There is a saying in aircraft design which asserts “if it looks good, it flies good”. The XP-92 Dart did not look particularly trustworthy. The cockpit position in particular would severely limit the pilot’s view, making it impossible to check his ‘six o’clock’ (i. e. directly aft). In addition the whole propulsion concept proved to be overly comphcated, with its large numbers and types of engines. Just one prototype was built with a conventional cockpit and only a jet engine. It was used primarily for research into delta-wing aerodynamics. In that role it did become the first US delta­wing aircraft (and the first of many delta-winged Convair fighter designs). When the test flights ended in late 1953 the mixed-propulsion interceptor concept was already considered outdated by the USAF.

Although Republic did not win the contest it was awarded a contract to build an experimental mixed-power interceptor based on its AP-31 design. This became the XF-91 Thunderceptor, a heavily modified version of the F-84F Thunderstreak with a liquid propellant rocket engine added for fast climb and high-speed interception. The original idea was to use a Curtiss-Wright XLR27 four-chamber rocket motor with a total thrust of 60,000 Newton but due to development problems a Reaction Motors XLR1 l-RM-9 with about half the power was mounted instead (the first XF – 91 flew with empty XLR27 housing pods and holes for two nozzles above and two below its jet exhaust). The XLR11 was the first US rocket engine developed specifically for use in aircraft, and was also used in the X-l and D-558-2 high-speed research aircraft described in the next chapter. The compact rocket engine was housed in a pod under the jet tail pipe and had four individual thrust chambers that burned ethyl alcohol and liquid oxygen. Each chamber had a fixed, non-throttleable thrust of 7,000 Newton but the total thrust of the cluster could be regulated by turning individual chambers on or off. It was fairly reliable.

The first XF-91 Thunderceptor [US Air Force],

The XF-91’s four-chamber XLR-11 rocket engine was placed directly under its jet engine’s exhaust pipe [US Air Force].

The two XF-91 prototypes retained the originally proposed and unusual inversely tapered wing, with the angle of attack relative to the fuselage being variable in flight between minus 2 and plus 6 degrees. The innovative wing provided excellent control at supersonic as well as at very low speeds, reducing the tendency of the tips to stall before the rest of the wing at low speeds (which can lead to a loss of control). The problem with limited space to house the main undercarriage was solved with a highly original arrangement of narrow twin wheels in a bicycle arrangement which retracted outwards into the broad wingtips. Otherwise the design was fairly conventional, with a cross-shaped tail and a regular cockpit with an ejection seat. The armament planned for its eventual operational service consisted of various types of cannon and air-to-air rockets. The prototypes were built and tested at Edwards Air Force Base,

California. The first flight was on 9 May 1949 with Republic’s chief test pilot Carl Bellinger at the controls, and it was powered only by the turbojet engine. It was 30 months before the rocket engine was delivered and fitted. On its maiden flight using dual-power on 9 December 1952, the Thunderceptor flown by Russel ‘Rusty’ Roth became the first US fighter (as it was intended to become an operational military combat aircraft) to fly faster than the speed of sound in level flight. It achieved Mach 1.07 powered by the jet engine with afterburner and two of the four rocket chambers. The aircraft later reached a maximum speed of 1,812 km per hour (1,126 miles per hour, Mach 1.71) on the combined thrust of the 24,000 Newton jet engine with afterburner and 27,000 Newton from the four rockets, which was very good for the time. With the XLR27 rocket engine it would probably have been able to reach Mach 2. It could climb to an altitude of 14.5 km (47,500 feet) in just 2.5 minutes and its maximum altitude was 16.8 km (55,000 feet). But the rocket equipment and propellant took so much space that the Thunderceptor could carry fuel to run the jet engine without afterburner for a mere 25 minutes, and even for that it had to be equipped with two large drop tanks with extra jet and rocket propellant. This would have restricted the plane to a point defense role.

The second XF-91 prototype was similar to its predecessor, although it was later retrofitted with a V-shaped ‘butterfly’ tail. It suffered a serious engine failure during take-off in the summer of 1951 with Bellinger at the controls, while famous test pilot Chuck Yeager flew chase in an F-86 Sabre (as described later in this book, in 1947 Yeager became the first man to break the sound barrier in the X-l rocket plane). In Yeager, an Autobiography, Carl Bellinger describes what happened: “Chuck radioed, ‘Man, you won’t believe what’s coming out of your engine.’ A moment later I got a fire warning light. ‘Christ,’ I said, T think I’m on fire.’ He replied, ‘Old buddy, I hate to tell you, but a piece of molten engine just shot out your exhaust, and you’d better do something quick.’ ” Bellinger was still too low to use his ejection seat and black smoke was filling the cockpit, bhnding him. Assisted by Yeager’s calm instructions Bellinger managed to land the Thunderceptor with its rear fuselage burning violently. He scrambled out and ran for his life. Only 90 seconds had elapsed between Yeager’s warning and the emergency landing. In spite of the serious damage, the aircraft was repaired and equipped with a V-shaped tail for further experiments.

Together the two XF-91s completed 192 test flights over the course of five years, but no more were built. As with the Soviet combined-propulsion interceptors, the XF-91 suffered from its short range and very limited total flight time, with the result that it was soon rendered obsolete by a new generation of powerful, more versatile turbojet-only fighters. The second prototype was scrapped but the first one (minus engines) is on display at the National Museum of the US Air Force near Dayton, Ohio.

Around the same time as the new mixed-propulsion aircraft, there were also plans to equip existing jet fighters with auxiliary rocket engines for increased performance for short periods. The Aerojet LR63-AJ-1 that burned white fuming nitric acid and JP-4 jet fuel (a 50-50 blend of kerosene and gasoline) and could produce a thrust of

23,0 Newton was under development for the F-84 Thunderstreak but during the Korean War the Air Force decided to use it in the F-86 Sabre because the early-type

Sabres were being outperformed by the Soviet MiG 15s in North Korean service and it was hoped that the rocket engine would enable the Sabre to counter the MiG threat. Versions of the engine were tried on an F-84G in 1952 and on an F-84F in 1954, but by the time the LR63 had been modified for the F-86 and flight tests on a prototype started in 1956 the Korean War was long over. Neither the mixed – propulsion version of the F-86 Sabre nor that of the F-84 Thunderstreak became operational.

The US Navy still saw some use for carrier-based mixed-power interceptors. In 1955 they ordered six North American FJ-4 Fury fighters with a Rocketdyne AR-1 rocket engine mounted above the standard jet engine exhaust pipe. The AR-1 ran on hydrogen peroxide and standard JP-4 jet fuel, taking the latter from the same tanks as fed the jet engine. The rocket gave the Fury an additional 22,000 Newton of thrust for short periods, sufficient to reach a speed of Mach 1.41 and a maximum altitude of 21.6 km (71,000 feet); considerably better than the normal limits of the standard FJ-4 at just over Mach 1 and 14.3 km (46,800 feet). Development of the rocket – boosted Fury, called the FJ-4F, became a top priority for a while after the US got into a bit of a panic concerning intelligence estimates for the high-altitude capabilities of the new Soviet Tupolev Tu-16 ‘Badger’ and the Myasishchev Mya-4 ‘Bison’ bombers. They urgently needed a naval high-altitude interceptor to thwart the new threat. The first two FJ-4Fs entered flight testing in April 1957 at the Naval Air Test Center, Patuxent River, Maryland, and achieved the aforementioned performance. Between them they made 103 flights, accumulating a total of 3.5 hours of rocket operation. However, the other four designated FJ-4s were never converted into mixed-power interceptors and, unsurprisingly, the FJ-4F concept was soon abandoned. Apart from the obsolescence of the rocket propelled interceptor idea, the risks involved with storing and handling hydrogen peroxide on warships put an end to the development.

Around the same time, in 1957 the Vought company was thinking about boosting their F8U Crusader naval fighter with a rocket engine. The idea was to fit a Reaction Motors XLF-40 rocket engine above the standard jet tail pipe in order to increase the rate of climb for high-altitude intercepts. The XLF-40 burned jet fuel and hydrogen peroxide and delivered 35,600 Newton of thrust. Unfortunately an engine exploded during a ground test and killed two Reaction Motors workers, prompting the engine company to withdraw from the project. Vought continued development for a while with the intention of switching to a Rocketdyne XLR54 rocket engine with a thrust of 26,000 Newton. Dummy engines were installed in two Crusaders to verify that the rocket engines would fit, but no real engines were ever installed. Vought had also planned to equip its new XF8U-3 Crusader III with a Rocketdyne AR-1 engine (as on the FJ-4F) but this was canceled together with the entire Crusader III project.

In the early 1960s the Navy introduced the North American Vigilante supersonic carrier-based bomber. The company planned to fit this aircraft with an auxiliary jet fuel/hydrogen peroxide rocket engine but this never happened. Neither was a mixed – propulsion interceptor version equipped with a Rocketdyne XLR46-NA-2 engine that they tried to sell to the Air Force as the Retaliator.

Back in Europe, France had entered the Second World War with obsolete aircraft that were quickly defeated by the Luftwaffe, and during the war there were no further developments. Afterwards, however, the country was determined to become a strong military power again and worked indigenously on advanced military aircraft. A large number of innovative test planes were built and flown during the late 1940s and the 1950s. In 1944 the government had created SEPR (Societe d’Etude de la Propulsion par Reaction) in order to develop its own rocket engine capabilities. The French do not appear to have had any interest in a pure rocket fighter like the Me 163 with its inherently Umited range and endurance. They were more interested in aircraft with combined rocket and turbojet propulsion. Studies for a rocket propelled interceptor with auxiliary jet engines began in 1948 at the aircraft manufacturer SNCASO (Societe Nationale des Constructions Aeronautiques du Sud-Ouest), commonly referred to as Sud-Ouest. The general idea was to use both jet and rocket engines for a rapid climb and high-altitude intercept, and jet engines alone for the return to base. The rocket engine was to be based on the SEPR design successfully employed by the Matra M04 missile and its successors. However, an engine on a missile has to work only once and for a short time; a rocket motor for a manned aircraft must be more reliable, reusable and maintainable. For operational readiness of the aircraft and engine, the rocket propellant would have to be readily storable (exit liquid oxygen with its extreme cooling requirements), fairly insensitive to changes in temperature and to impurities in the tanks (exit hydrogen peroxide) and easy to produce and handle (again exit hydrogen peroxide). Somewhat surprisingly, it was decided to use nitric acid as oxidizer even although the corrosion of the engine and aircraft structure must have been apparent. (As we have seen, this problem made the Russians eventually discard nitric acid as a propellant for rocket interceptor aircraft.) The fuel was ‘furaline’, a mixture of 41% furfuryl-alcohol, 41% xylidene (dimethylaminobenzene) and 18% methanol; also a rather surprising choice since it was fairly difficult to produce, especially in comparison with the low-grade kerosene used in the Russian rocket engines for aircraft. An advantage over kerosene was that furaline is hypergolic with nitric acid, eliminating the need for an ignition system.

As the concept of rocket thrust (let alone mixed-propulsion) on a manned aircraft was totally new to France, it was decided first to build a relatively simple proof-of – concept design before embarking on the expensive and risky development of a high – performance aircraft. Therefore one of the existing SNCASO SO-6020 ‘Espadon’ (Swordfish) jet prototypes was fitted with a SEPR 25 engine that ran on nitric acid and furaline fed from tanks on the wingtips, and renamed the SO-6025 Espadon. The light alloy engine was slung under the rear fuselage and produced a modest thrust of 1,500 Newton. It was regeneratively cooled by a flow of nitric acid running around the chamber and nozzle before being injected into and burned in the combustion chamber. The pumps supplying the propellants were powered by the aircraft’s jet engine, the method pioneered by BMW during the war in their 003R combined jet/ rocket engine. The modified Espadon was intended to demonstrate the safety and reliability of the rocket engine and the capabilities of the mixed-propulsion concept. The first ground test of the plane fitted with the rocket engine was in May 1951 but

The Espadon under rocket power.

the first flight did not occur until 10 June 1952. A total of 76 rocket propelled flights had been made by July 1955, with the rocket-equipped Espadon becoming the first European aircraft to reach Mach 1 in level flight on 15 December 1953. An improved engine, the SEPR 251, was installed in another Espadon and flown 13 times between March 1953 and October 1954.

Based on the successful trials with the two Espadons, the French Air Force invited proposals from French aircraft manufacturers to design a Ughtweight interceptor with jet engines, rocket engines, or both, which would be able to achieve Mach 1.3, have a high climb rate and be able to operate from rough airfields.

Nord Aviation proposed the Nord 5000 ‘Harpon’ (Harpoon), which was initially conceived as a rocket propelled interceptor with sharp swept-back canard stabilizers and double-delta wings (in which the sweep angle of the leading edge changes at a certain point, as on the Space Shuttle). Morane-Saulnier proposed the MS-1000, of which not much is known. Neither was selected for further development. The Nord 5000 never got further than a wooden scale model.

SNCASO won a contract with designer Lucien Servanty’s SO-9000 Trident. This fast looking, bullet-shaped supersonic interceptor had a rocket engine fitted in its tail, augmented by a pair of small Turbomeca Marbore II jet engines with 4,000 Newton of thrust each for a fuel-economic return flight after the rapid climb and intercept on combined propulsion. Its wings were straight but relatively thin to minimize drag at high speeds, and had an overall span of 7.57 meters (24.8 feet). The light jet engines were on the wing tips. This kept the fuselage aerodynamically clean but required the wings to be especially strong and stiff, and hence rather heavy. The SEPR 481 rocket engine had three combustion chambers and these could be ignited and extinguished individually to give it a throttling capability in three steps of 15,000

SO-9000 Trident [SNCASO].

Newton and an impressive maximum thrust at sea level of 45,000 Newton. At an altitude of 11 km (36,000 feet) the engine even produced 52,000 Newton owing to the lower external air pressure trying to push the rocket exhaust back in. The gas generator driving the turbopumps ran on the same propellants as the combustion chambers, meaning that the rocket engine did not rely on power from the jet engines; it was an independent unit. One complication was that a mixture of water and methanol had to be injected into the gas generator to reduce the gas temperatures and prevent the turbine blades of the turbopump from melting.

To enable the pilot to escape from a stricken plane at high speed, the entire nose section of the Trident could be jettisoned; a system which offered the pilot a lot of protection but was a much heavier solution than an ejection seat (a similar system had already been used in the Heinkel 176 rocket plane and was also incorporated in the Douglas D-558-1 Skystreak, D-558-2 Skyrocket and Bell X-2 research aircraft, and similar concepts would later be suggested as possible methods for astronauts to escape from malfunctioning shuttles and spaceplanes). The control system was well suited to transonic and supersonic flight speeds, as all three tail surfaces (the vertical and two horizontal stabilizers) were all-moving, rather than small separate rudders and elevators. This prevented the controls from locking up due to Mach shock waves forming at the control hinge lines (the aforementioned dangerous ‘shock stall’). At low speeds normal ailerons on the wings were used, but at high speeds these would be locked in order to prevent shocks forming and in this configuration roll would be controlled by differential use of the horizontal tail surfaces (which would thus act as elevons). The tricycle undercarriage was retracted into the fuselage. The fully-loaded weight was 5,055 kg (11,140 pounds); more than the original 4,000 kg requirement but still relatively low for an interceptor.

SNCASO built two Trident prototypes. The rocket engine was not fitted for the initial test flights, as the engineers first wanted to determine how the plane handled at low speeds under jet power alone. The first jet-only prototype took off from Melun – Villaroche on its maiden flight on 2 March 1953 with test pilot Jacques Guignard at the controls. After completing a number of other successful flights the prototype was revealed to the public at the famous international air show at Le Bourget near Paris in June. Guignard also took the second Trident (still lacking a rocket engine) for its first flight on 1 September 1953 but owing to trouble with the relatively weak jet engines he was unable to gain sufficient height after take-off and hit an electricity pylon. The plane struck the ground so violently that the escape module broke off. Guignard was badly injured but survived. To preclude further take-off accidents the one remaining Trident was refitted with two licence-built early versions of the British Armstrong Siddeley Viper jet engine that had almost twice the thrust of the previous Marbore II engine. Test pilot Charles Goujon resumed the test flights on 16 January 1954. Later in the autumn the plane and the rocket engine were deemed to be ready and the first rocket propelled flight using one chamber of the SEPR 481 rocket engine was made on 9 September. Over the next 2 years the Trident made a total of 24 rocket propelled flights to explore its performance. It was capable of exceeding Mach 1 on jet power alone when pushed into a shallow dive but with the SEPR 481 it could achieve a top speed of Mach 1.63 and an altitude of nearly 15 km (52,000 feet). After the flight test program was terminated in April 1956 the first and only surviving Trident prototype was retired to the Le Bourget Musee de l’Air et de l’Espace museum. It is still there, suspended from the ceiling flying at Mach 0 but nevertheless still looking fast and beautiful in its bare metal color scheme with French national markings. The wingtip-mounted jet engines do give it a decidedly 1950s’ look, though.

The success of the Trident test program prompted the French Air Force to place an order for an improved version to serve as an operational mixed-power

SO-9000 Trident in the museum at Le Bourget [Mikael Restoux].

SO-9050 Trident II [SNCASO],

interceptor. Designated the SO-9050 Trident II, this plane was equipped with the new SEPR 631 two-chamber rocket engine with each chamber delivering 15,000 Newton of thrust (although less than the three-chamber engine this was a more propellant-economical compromise between climb rate and endurance). The fully – laden take-off weight was about 5,900 kg (13,000 pounds) and the wingspan was 6.95 meters (22.8 feet). The Trident II was still fitted with the Viper engines of British origin but the detachable cockpit concept of the SO-9000 was abandoned in favor of an ejection seat. In order to be able to mount air-to-air missiles under the fuselage the undercarriage legs were lengthened.

Flight testing of the first Trident II began on 19 July 1955 powered by the jet engines only. The first rocket powered flight was on 21 December. The second plane was damaged beyond repair shortly thereafter when both jet engines flamed out soon after take-off and test pilot Guignard had to make an emergency crash landing. This time he was only slightly injured. A third prototype, equipped to be flown without a pilot, controlled from the ground by radio link, was used as a replacement for the lost aircraft (the remote control feature was never actually used). The first prototype was used to explore the highest performance of the design while the third was for tests at low and medium speeds. The first Trident II disintegrated in mid-air during a dive on 21 May 1957 killing pilot Charles Goujon. The real cause for this disaster was never discovered but an explosion due to propellant leakages and structural failure induced by the abnormal aerodynamic loads were high on the list. However, confidence remained high that the Trident II would enter operational service. The magazine Flight reported that at the Le Bourget air show, which opened only 3 days after the crash, “a fascinating large-scale model shows a hypothetical Trident II base with camouflaged dispersals, refueling and servicing points, missile protection and other modem conveniences”. The surviving prototype was quickly brought up to the standard of its predecessors so that high-performance test flights could continue. It eventually reached a top speed of Mach 1.95 and a record altitude of 24 km (79,000 feet), both very impressive for the time. Jacques Guignard set two time-to-altitude records using this aircraft: 2 minutes and 36 seconds to 15 km (49,000 feet) altitude, and also 3 minutes and 17 seconds to 18 km (59,000 feet).

Despite losing two aircraft during the test campaign (not unusual for experimental planes in those days), in June 1956 the government placed a pre­production order for the operational Trident III. This had a Turbomeca Gabizo jet engine at the tip of each wing, each developing 11,000 Newton of thrust on its afterburner. The rocket main engine was the SEPR 632, which was very similar to the 631 except for using ‘tonka’ fuel (an approximately 50-50 mix of triethylamine and xylidine which is hypergolic with the nitric acid oxidizer) rather than furaline, which was difficult to produce. The new aircraft was to be equipped with a fire control radar and Matra R 511 air-to-air missiles. However, the first three production aircraft had the 631 engine and they did not carry armament. These early planes were tested until October 1958, with Captain Jean Pierre Rozier reaching an altitude of 26 km (85,000 feet), although this was not registered as a record because the flight was not supervised by the FAI (Federation Aeronautique Internationale), the international society responsible for logging world aviation records. The top speed set with a Trident III was Mach 1.95 at an altitude of 22.1 km (72,500 feet). But the whole program had been canceled in April 1958 with only four aircraft completed, another two almost finished, and four more in various stages of construction. Unfortunately all of the Trident IIs and Ills were scrapped, leaving us not a single example of this impressive product of the rebirth of French aeronautical prowess. This is all the more lamentable because the Trident III set an official FAI altitude record for a completely self-propelled rocket plane of 24,217 meters (79,452 feet) on 2 May 1958, achieved by pilot Roger Carpentier in response to the project’s cancellation a week earlier. The Trident III actually had the potential to push to 30 km (100,000 feet). Other experimental rocket aircraft had already gone higher than that but, unlike the Trident III, they were air-launched by carrier planes and therefore could not claim official FAI records for self-propelled flight.

As the Russians and Americans had found out, mixed-propulsion planes based on small jet engines and big rocket engines had much too little range and endurance for practical military use. At full rocket power the Trident III could stay in the air only for about 4 minutes! Like other air forces, by the end of the 1950s the French shifted their interest away from short-range interceptors to multi-purpose long-range fighter aircraft. In the future, interceptors would be powered by large jet engines and exploit rocket power only for extra boost. Following this philosophy, aircraft manufacturer SNCASE (Societe Nationale des Constructions Aeronautiques du Sud-Est, not to be confused with SNCASO, although in 1957 they merged to form Sud Aviation) won a contract to develop the SE-212 Durandal, a delta-wing mixed-propulsion interceptor named for a mythical medieval French sword. Developed by a team headed by Pierre Satre, it was primarily powered by a SNECMA Atar 10IF jet engine which delivered a thrust of 43,000 Newton with afterburner, and had a single-chamber 7,500 Newton SEPR 65 auxiliary rocket engine. With a total weight of 6,700 kg (15,000 pounds) it

The second Durandal at the 1957 Paris Air Show.

was a bit heavier than the Trident II and III. The Durandal was intended to be armed with a single AA-20 air-to-air missile carried under the centerline or two 30­mm cannon. As with the Espadon, the pumps of the rocket engine were driven by the jet engine through an auxiliary gearbox. The rocket engine ran on the hypergolic combination of nitric acid and triethylamine-xylidine (TX).

Development of the SEPR 65 started in December 1953 and its first ground test was performed on 4 November 1954. The initial Durandal prototype took to the sky on 20 April 1956 and its first rocket propelled flight was on 19 December that year. The second prototype first flew on 30 March 1957 and it employed rocket power for the first time in April 1957. During its test program the Durandal reached a speed of 1,444 km per hour (898 miles per hour, Mach 1.36) on jet power alone and 1,667 km per hour (1,036 miles per hour, Mach 1.57) with rocket thrust, both at an altitude of about 12 km (39,000 feet). The two aircraft made a total of 45 rocket flights before the program was terminated in May 1957. The main reason for the cancellation was the fact that carrying a single missile it would have only one chance to hit its target and would then be unable to defend itself. This gave it little advantage over a guided surface-to-air missile, and so rendered it of very limited use to the French Air Force. Parts of the first Durandal are in storage at the Musee de l’Air et de l’Espace at Le Bourget, near Paris. It was intended to fit a derivative of the SEPR 65 engine, the SEPR 651, to the small Nord 1405 Gerfaut II delta-wing experimental aircraft but this was never done.

Aircraft company Dassault also developed a small mixed-propulsion interceptor, the delta-winged MD-550 – Mystere Delta 550. It had no horizontal stabilizers but an exceptionally large vertical stabilizer. It was to have two Turbomeca Gabizo engines side-by-side for jet propulsion and a single-chamber SEPR 66 rocket engine located between and beneath them. The rocket had a gas generator driving the turbopumps and burned nitric acid and furaline to deliver a thrust of 15,000 Newton. After the development started in September 1953 the first SEPR 66 ground tests took place on 10 January 1955. Over a year later, on 19 January 1956, the engine made its first test flight on the Mystere IV B05, a ‘flying test stand’ based on the Dassault Mystere IV fighter-bomber. On the tenth rocket flight an explosion of the combustion chamber due to an ignition delay severely damaged the tail of that aircraft. Nevertheless, the pilot managed to land. The problem was quickly traced to the very low temperature at the high altitude at which the plane was flying, and rapidly resolved. At the end of this testing the Mystere IV B05 had made 55 rocket flights and proven that the concept worked. A prototype of the MD-550 (lacking a rocket engine) was built and flown for the first time in June 1955 on jet power provided by two temporary Viper engines (which were, however, never replaced with the intended Gabizo turbojets). This aircraft was then modified, equipped with a smaller tail fin, afterburners, and a SEPR 66, and named the Mirage I. Pilot Gerard Muselli ignited the SEPR 66 for the first time in flight on 17 December 1956, pushing the Mirage I up to an altitude of 12 km (39,000 feet) and a speed of Mach 1.6. Previously, the maximum speed attained on jet power alone had been Mach 1.2 so the rocket thrust added a considerable 0.4 Mach. But the project was terminated in late January 1957 with the aircraft having made only five rocket flights. Like the SNCASE Durandal (canceled 4 months later) the Mirage I had very limited military use because it could only carry a single air-to-air missile.

In reaction to the Air Force’s shift in interest to multi-purpose fighters with a long range and multiple armament, Dassault designed the Mirage II. This was a bigger aircraft powered by two Gabizo engines and a SEPR 661, which was a dual-chamber version of the SEPR 66 with the same 15,000 Newton maximum thrust but the option of a 7,500 Newton intermediate thrust by using just one chamber. The SEPR 661 was only ever fired on a ground test bench and the single Mirage II prototype only flew on jet power, but this aircraft design formed the basis for the subsequent, much more successful mixed-power Mirage III.

The Mirage III was a very slender delta-wing interceptor based on the innovative ‘area ruling’ concept which dictates that to limit aerodynamic drag, changes to the frontal cross section of an aircraft should be made as gradual as possible. This in turn means that near the wings, the fuselage must become thinner in order to maintain a constant cross section with respect to the nose and tail. This gave rise to the famous ‘wasp waist’ (or ‘Cola Cola’ or ‘Marilyn Monroe’) configuration that can be seen on many supersonic fighters. With its maximum take-off weight of 13,500 kg (29,700 pounds) it was much larger and heavier than the Tridents and Durandal. The Mirage III was optimized as a high-altitude interceptor, emphasizing rate of climb and speed rather than maneuverability and low-speed handling. Its primary propulsion was a single SNECMA Atar 101G.1 turbojet which delivered a thrust of 44,000 Newton with afterburner. The pumps of its SEPR 84 rocket engine were powered by the jet engine, feeding the motor with nitric acid and TX. The combustion chamber had a double wall through which nitric acid flowed en route to the chamber, thereby cooling the engine. Another smart feature was that the high – pressure kerosene also served as a hydraulic fluid for operating various valves. Even although it had only one chamber, the SEPR 84 could provide a maximum of 15,000 Newton as well as an intermediate level of 7,500 Newton of thrust. The development of the new rocket engine started at the end of 1956, and it was first ignited in the air on 12 July 1957 on the Mirage III 001 prototype piloted by Roland Glavany. This plane flew a total of 30 missions, during which Glavany managed to reach Mach 1.8 in level flight at an altitude of 12 km (39,000 feet).

The success of the prototype led to a pre-production order for ten Mirage IIIAs, which were intended to be multipurpose fighters rather than merely interceptors. The Mirage IIIA was longer than the prototype, had a larger wing, and was fitted with the powerful Atar 9B turbojet which could dehver a thrust of 58,900 Newton with afterburner. The Mirage IIIA 02 prototype was initially fitted with the first improved pre-production SEPR 840 engine and had sufficient propellant for an 80 second burst of thrust. The aircraft made its first rocket propelled flight on 18 February 1958, and it performed a total of 15 rocket flights with the SEPR 840. It went on to make an impressive 444 rocket propelled test flights with the operational SEPR 841 version of the engine. On 24 October 1958 Glavany nudged the Mirage IIIA 01 prototype past Mach 2 in level flight with the help of a SEPR 840, becoming the first European pilot to do so. On 25 January 1960 a rocket boost enabled the aircraft to reach an altitude of 25.5 km (83,700 feet). The Mirage IIIA design suffered from corrosion of the nitric acid tank and damage to the underside of the fuselage caused by divergence of the rocket exhaust at high altitude (where the lower air pressure permits the exhaust to expand into a wide plume). However, these problems were solved by changing the material of the acid tank and beefing up the fuselage thermal protection.

The first major production version, the Mirage IIIC, had its maiden flight in 1960 and entered operational service in December the following year (the IIIB was a two- seat trainer and had no provision for a rocket engine). It was armed with two 30-mm cannon and had five attachment points for missiles, bombs or extra fuel tanks. The SEPR 841 with its propellant tanks could be added as a separate package to the aft fuselage. If the rocket motor was not needed for the mission, it could be left vacant or replaced by an additional jet fuel tank. The Mirage IIIC remained in use with the French Air Force until 1970, making 1,505 rocket propelled flights involving 2,064 ignitions of the SEPR 841. The rocket engine could give a very powerful kick which allowed the aircraft to accelerate from Mach 1 to Mach 2 in only 90 seconds; on jet power alone this would take about 4 minutes. In terms of added engine weight, this boost came fairly cheaply. The Atar 9B-3 turbojet had a weight of 1,460 kg (3,210 pounds) and on afterburner could deliver 58,900 Newton (equivalent to 5,890 kg) of thrust; a thrust to weight ratio of 4. The SEPR 841 had a weight of only 158 kg (348 pounds) and delivered 15,000 Newton (equivalent to 1,500 kg) of thrust: a thrust to weight ratio of 9.5; more than double that of the jet engine. Operating the rocket engine from the cockpit was also very simple: the instrumentation was limited to an ignition switch, a handle for thrust level selection, a control light to indicate that the engine was operating and another light to indicate a malfunction.

The ensuing Mirage HIE, which had various improvements to better suit the role of ground attack aircraft, could be fitted with a SEPR 844 engine in which the TX

Ground testing of an SEPR 844 on a Swiss Mirage IIIS [F +W Emmen].

fuel of the 841 was replaced by kerosene (the HID was the two-seat trainer version of the HIE). The main benefit of this was that the rocket engine could get its fuel from the same supply as the jet engine. But because this new propellant combination was not hypergolic, a small tank of TX still had to be carried to ignite the engine (the tank was sufficient for two in-flight starts). The rocket engine could operate for a

A SEPR 844 on a Swiss Mirage IIIS operating in flight [F +W Emmen],

total of 2 minutes and could be stopped and reignited in flight. The SEPR 844 could deliver a thrust of 12,600 Newton at sea level and 15,600 Newton at an altitude of 20 km (66,000 feet) for a specific impulse of 220 seconds. The Mirage HIE was also built under licence in Switzerland by F + W Emmen as the Mirage IIIS for the Swiss Air Force, which opted to equip its aircraft with SEPR 844 rocket engines.

The Mirage III went on to become the most successful operational mixed-power fighter aircraft ever by logging over 20,000 rocket propelled flights. Despite the use of nitric acid, a spent rocket pack could be made ready for a new mission in less than 15 minutes. This demonstrated that rocket engines could be used by an operational fast – reaction interceptor aircraft. Acid corrosion problems such as experienced by the Russians were avoided by adding a corrosion inhibiter to the nitric acid propellant. A total of 164 SEPR 841 and 111 SEPR 844 engines were built and each could be used for an average of 35 flights. SEPR rocket packs were used on Mirage III fighters by the air forces of no fewer than six countries: France, Libya, Pakistan, South Africa, Spain and Switzerland. This made the SEPR 84 family the most built, most reused, and most operational man-rated (i. e. intended for use on piloted aircraft) series of rocket engines ever. Nevertheless, improved jet engines, accurate long-distance air-to – air missiles, and ground radar networks capable of providing advanced warning of airspace intruders meant that the auxiliary rocket engine was no longer needed, and subsequent French fighter aircraft did not have an add-on rocket capability (although some could be fitted with RATO boosters for take-off). France ended the operational service of the SEPR 844 in 1984 but the Swiss continued to use them until 1996.

Hence the French interceptor philosophy had moved from the concept of a rocket plane with auxiliary jet engines, to a turbojet aircraft with auxiliary rocket power, to a pure jet fighter without any provision for in-flight rocket propulsion. It meant the end for rocket propelled military aircraft in France.

The United Kingdom, although at the forefront of turbojet engine technology, had not done much on rocket and rocket plane development during the war. Based on the knowledge and equipment they managed to capture in Germany in 1945, and assisted by several German rocket scientists, Britain rapidly caught up on rocketry. The initial focus was on the development of jettisonable RATO units, most notably the Lizzie liquid oxygen/petrol engine for the Vickers Wellington bomber, a modified version of the German RI-202 for the Avro Lancastrian (a passenger aircraft based on the Lancaster bomber), and the ‘cold’, compressed-air-fed hydrogen peroxide/ calcium permanganate Sprite engine for the de Havilland Comet jet airliner.

The Sprite, developed in-house by the aircraft company, was meant to help the airliner take off from the many ‘hot and high’ airports in the British Empire where the air density, and therefore the aircraft’s lift at take-off, were diminished by high temperatures and high elevations. The two Sprites on the Comet could each provide a 22,000 Newton boost for 16 seconds. Flight tests on the aircraft prototype started in May 1951 but the system was never used for real operations; injecting water into the combustion chambers of the jet engines proved to be a simpler and cheaper method of gaining the necessary extra boost.

An important innovation for an upgraded Sprite was the use of permanent silver plated nickel gauze packs as the catalyst, replacing the expendable liquid potassium permanganate solution used in the Walter engines as well as the early version of the Sprite. This type of catalyst, which was not consumed and hence did not need to be replenished for each flight, would be incorporated in all following British hydrogen peroxide rocket engines. The Sprite later became the Super Sprite, in which kerosene was injected into the combustion chamber and then spontaneously ignited due to the high temperature of the steam and oxygen gas developed by the decomposition of hydrogen peroxide (which the British named High Test Peroxide, HTP). At 19,000 Newton the engine’s thrust was actually less than that of its predecessor, but it had an increased bum time of 40 seconds. The Super Sprite was used operationally by the Vickers Valiant bombers of the Royal Air Force in the early 1950s.

In early 1951 the estimated capabilities of near-future Soviet long-range bombers led the British to believe they would need to be able to intercept bombers flying at extreme altitudes as well as at speeds up to Mach 2. The limitations of the available early warning radar system, in combination with the expected large scale of an air attack, meant that the United Kingdom required a large force of high-performance interceptor aircraft with very fast climb rates and high service ceilings. However, the existing British fighters were nowhere near capable of thwarting the perceived Soviet threat. Based on their experience with RATO rocket engines, the British made plans to use rocket engines for new high-speed, high-altitude manned interceptors.

Initially there was also some interest in pure rocket propelled interceptors, with a government specification leading to the Short P. D.7, Bristol 178 and Hawker P.1089 concepts. The Short P. D.7 design had large delta wings that provided room for four large internal kerosene tanks, while the fuselage housed the additional liquid oxygen needed for the Screamer rocket engine (more on this engine later). The high-subsonic airplane was to have been able to reach an altitude of 9 km (30,000 feet) in less then 3 minutes from the moment that its wheels started to roll over the runway. But once at the altitude of enemy bombers the aircraft would have hardly any propellant left, making it difficult to engage anything. This very limited range meant that the rocket interceptors would have to be stationed as close as possible to their potential targets and in very large numbers, which translated into a large number of individual bases along the coast. Like the French, the British quickly found range and endurance of a pure rocket fighter too limited to be of much use, and hence regarded rocket power on aircraft as being practical only if combined with turbojet thrust. They therefore asked their industries to propose concepts for interceptors using both jet and rocket engines. In addition, they wanted competitive industries to integrate rocket engines based on different propellants in their designs, in order to preclude becoming stuck in dead-end developments.

For the Royal Navy, the Hawker company proposed a new version of their Sea Hawk carrier-based jet fighter augmented with an Armstrong Siddeley Snarler rocket engine running on liquid oxygen and water-methanol (i. e. methyl-alcohol – water) and producing a thrust of 9,000 Newton for an increased rate of climb. The engine was installed in the P.1040 prototype of the Sea Hawk, which was fairly straightforward because the double exhausts of the single Rolls Royce Nene turbojet engine were positioned at the sides of the fuselage, thus leaving the tail conveniently free for the rocket. The fuselage was reinforced to deal with the extra rocket thrust, the Snarler’s turbopumps were connected to the 23,000 Newton jet engine via a gearbox, and the jet fuel tank capacity was decreased in order to make room for the new liquid oxygen and water-methanol tanks. There was enough propellant to run the rocket engine for 2 minutes and 45 seconds at full power. The aircraft was subsequently redesignated P.1072.

Design of the Hawker P.1072.

The plane’s Snarler engine was first ignited during a flight on 20 November 1950 by Hawker’s chief test pilot Trevor Wade. The August 1954 edition of the magazine Flight described it as follows: “The day was dull and overcast, and the P.1072 could be heard only distantly through dense cloud. Those on Bitteswell airfield were quite ignorant of whether Wade had fired the rocket or not, and were still waiting expectantly when the fighter appeared low down at full power. Passing overhead, Wade did a smart roll; all had gone well.” Six rocket propelled flights were made in total, three piloted by Wade and three by Neville Duke. The rocket engine was only fired during a steep climb in order to preclude the plane developing so much speed as to enter the transonic regime, for which its aerodynamics were inappropriate. With rocket thrust the plane could (theoretically) reach an altitude of 15 km (50,000 feet) within 3.5 minutes of its wheels starting to roll over the runway. This translated into an impressive rate of climb of over 73 meters per second (240 feet per second). But this performance could never be verified because the absence of a pressurized cabin meant that the pilot could not fly higher than 12 km (40,000 feet). Nevertheless the experimental prototype did establish that the concept of an auxiliary rocket engine worked. The test campaign also established that despite the cryogenic liquid oxygen the aircraft could be kept fully fueled for extended times without the engine’s valves freezing up (although the oxygen would boil off and so require constant topping up). Interestingly, the aforementioned Flight article does not comment on the fact that the Snarler engine exploded when Duke attempted a restart during the sixth flight on 19 January 1951; presumably that was still secret information at the time. This incident is believed to have been caused by a small quantity of fuel left in the system after the previous firing of the engine, which then instantly exploded upon reignition. Luckily the damage was relatively small, and the pilot landed the plane safely. The P.1072’s tail section was repaired but by then the Navy had decided that using afterburners on their jet engines was an effective and less complicated means of increasing thrust on fighter aircraft. The rocket propelled Sea Hawk was thus canceled. The P.1072 never flew again, and unfortunately it does not exist anymore.

The Avro company responded to the government’s request for a mixed-power interceptor with a proposal for an aircraft based on the Armstrong Siddeley Viper 2 turbojet which produced a sea level thrust of 7,000 Newton, in combination with an Armstrong Siddeley successor to the Snarler rocket engine with the equally inspired name Screamer. The Screamer burned liquid oxygen and standard aviation kerosene, produced a maximum thrust of 35,000 Newton, and could be restarted in flight. The Snarler had used water-methanol as fuel because of its good cooling characteristics. The Screamer could run on the same kerosene fuel as an aircraft’s jet engine, which meant it did not require separate fuel tanks, but the disadvantage was that kerosene is a much less effective coolant than water-methanol. Hence the Screamer depended on a water-cooled combustion chamber and nozzle. In comparison to an engine based on hydrogen peroxide the Snarler’s and especially the Screamer’s propellant was less costly and safer, but the liquid oxygen had the disadvantage of requiring very cold storage. Unlike the Snarler, the Screamer engine was required to run independently of the jet engines, which could therefore be turned off. It had its own gas generator in which a small amount of liquid oxygen and kerosene were burned to provide gas to drive the turbopumps to feed propellant to the rocket engine’s combustion chamber (water was injected into the gas generator’s exhaust to cool the very hot gas in order to prevent it damaging the turbopump assembly).

The resulting Avro 720 aircraft was a tailless delta-wing design somewhat similar to the later French Mirage. It was to be primarily constructed of metal honeycomb sandwich, a light but strong material (although difficult to repair) now widely used for spacecraft. It was intended to arm it with a pair of de Havilland Blue Jay (later renamed Firestreak) passive infrared air-to-air missiles carried on underwing pylons. The Type 720 was meant for the Royal Air Force but a derivative called the Type 728 would satisfy the specific requirements of the Royal Navy (it would be equipped with the much more powerful Gyron Junior jet engine instead of the Viper, and have a strengthened undercarriage for safe operation from aircraft carriers). Thanks to the low drag, exceptionally low weight and good altitude performance of the Screamer, both versions were to be able to reach Mach 2, be able to climb to 12 km (40,000 feet) in only 1 minute 50 seconds and have a maximum altitude of 18 km (60,000 feet). Armed with two heat seeking missiles under the wings the 720/728 was expected to be fully capable of intercepting high – performance bombers and blowing them out of the sky. A contract was issued for the construction of two prototypes and a structural model for ground tests. In 1956 the first aircraft was almost finished when the entire project, including the development of the Screamer, was canceled. The military had somewhat belatedly come to the conclusion that using liquid oxygen in an interceptor aircraft was not such a good idea after all. Oxygen has to be kept at the very low temperature of minus 183 degrees Celsius (minus 279 degrees Fahrenheit) in order to keep it liquid. Even with extensive insulation around the aircraft’s internal tanks it was expected that onboard oxygen would boil off at a rate of 4% per hour. So either it must be stored in special tanks and pumped into the aircraft shortly before flight, or if it is stored in tanks in the plane it must be topped off. As already determined by the Germans in 1939 when they reviewed von Braun’s early design for a vertical take­off rocket interceptor, this is a major problem for a quick-reaction interceptor that must always be ready for a rapid take-off.

Despite the cancellation of the Avro 720/728, a scheduled test of the Screamer on a Gloster F8 Meteor jet fighter continued. The rocket engine was to be placed on the underside of the aircraft, along with a liquid oxygen tank that would be jettisoned if the pilot ever had to make an emergency belly landing. The rocket’s fuel and water tanks were inside the aircraft’s fuselage. Ground firing tests were performed, but the full modification of the Meteor was not finished and its Screamer was never used in flight.

With Avro also out of the competition, Saunders Roe was the only company left working on a true, newly-designed mixed-power interceptor. Their SR.53 developed under the leadership of chief designer Maurice Brennan was a single-seat delta-wing aircraft which combined the de Havilland Spectre rocket with an auxiliary turbojet. The rocket engine would be used for rapid ascent and attack, while the jet engine mounted above it would enable the aircraft to make a low-power return to base. Like the Avro Type 720, the SR.53 would only be armed with two de Havilland Blue Jay

Design of the Saunders Roe SR.53.

(Firestreak) missiles, in this case mounted on the wingtips. The SR.53 was equipped with a T-tail so that the horizontal stabilizers were kept out of the wake of the wings. These stabilizers were of the all-moving type, thus avoiding the shock-stall problems of conventional elevators. Both the wings and the stabilizers were kept thin in order to limit high-speed drag and had a thickness ratio of only 6%. The take-off weight of 8,400 kg (18,400 pounds) included 5,000 kg (11,000 pounds) of propellant.

The de Havilland Spectre rocket engine was specifically developed for a mixed- propulsion interceptor and was heavily based on wartime German research by the Walter company. Hydrogen peroxide was decomposed by being passed through the silver plated nickel gauze catalyst and the resulting hot oxygen used to bum standard aviation kerosene. Cooling of the combustion chamber was achieved by running the hydrogen peroxide through a double wall. Pumps driven by the SR.53’s small jet engine fed the propellants to the Spectre’s combustion chamber from tanks that were pressurized with air. This pressurization served to suppress cavitation (the forming of bubbles due to local low pressure) in the pumps. Static firings of the rocket engine began in 1952 and demonstrated that it could be throttled from 8.9 to a maximum of

36,0 Newton. The Spectre was flown on a modified Canberra jet bomber test bed (described later) in 1957 before it was deemed ready for incorporation in the SR.53. The auxiliary jet engine was the Armstrong Siddeley Viper 8, which could produce a maximum thmst of 7,300 Newton. This also powered the onboard systems. It was not equipped with an afterburner, and in fact did not provide enough power to enable the plane to fly around after an overshoot on landing (as with the German wartime Komet, the pilot was committed to his approach). The rocket and jet engine received their kerosene from separate tanks but if necessary the turbojet could draw upon the rocket engine’s kerosene wing tanks. The combined power of both engines gave the SR. 53 a top speed of Mach 2.2 and a maximum altitude of 20.4 km (67,000 feet). From standing on the runway, it could climb to an altitude of 15 km (50,000 feet) in just 2 minutes and 12 seconds with a very impressive climb rate of 270 meters per second (890 feet per second). However, it had sufficient propellant on board only for 7.5 minutes of full rocket thrust.

The US backed the development financially with $1.5 million (equivalent to $12 million in 2010) and the UK Ministry of Defence signed a contract with the company in May 1953 to build three SR.53 prototypes, although the third one was canceled in January 1954 to cut costs. The Ministry of Defence had expected the first SR.53 to fly in mid-1955 but problems with the Spectre and its propellant supply system (one engine even exploded in a ground test) delayed delivery of the first SR.53 prototype to January 1957. With registration number XD 145, this machine was sent by water (the factory was on the Isle of Wight) and by road to the Ministry’s aircraft testing site at Boscombe Down for ground test firings of its Spectre engine. The aircraft took to the skies under rocket power for the first time on 16 May 1957 with test pilot John Booth at the controls. It was found to behave as expected with good flight characteristics in general, although the somewhat lower than advertised performance of the Viper engine in combination with an increase in aircraft weight meant that its performance fell short of the specification.

A potential problem with the aircraft’s recovery from a so-called ‘deep stall’ was discovered when a Gloster Javelin jet fighter, which had a similar configuration with

The SR.53 at the Famborough air show in 1957.

delta wings and T-shaped tail, crashed during a test flight because at extreme angles of attack the turbulent wake of the stalled wings blanketed the horizontal stabilizer, rendering its elevators ineffective and thereby making a recovery impossible (as this involved using the elevators to push the nose down). To prevent this from happening to the SR.53 a set of four small solid propellant rockets were installed in the rocket engine’s cowhng in the tail: if the pilot ever discovered himself in a deep stall, he could ignite these motors to pitch the aircraft back down (the motors were units that the Royal Marines used to shoot hooks and lines up for cliff assaults). There was also an issue with the Viper flaming out when the SR.53 was pushed over after a steep climb, resulting in a short duration of zero G (the principle on which aircraft for 0-G weightlessness-simulating research are based) that disrupted the flow of propellants. More important, however, was a serious design flaw that was discovered during the test flight campaign: the rear airframe structure suffered badly from the high level of acoustic energy (i. e. sound vibrations) generated by the powerful Spectre engine. The loads were not sufficient to immediately damage the aircraft but prolonged firing of the engine rapidly weakened the structure. This is called ‘fatigue’ and is a well known issue in aircraft design and maintenance (the problem is similar to how a spoon weakens when you bend it several times until it eventually breaks due to the growth of micro-fractures in the metal). Fatigue on the rear structure of the aircraft delayed the development program even further.

Meanwhile the Royal Air Force had come to the conclusion that it really required something better than the SR.53, which was hampered by its weak turbojet and very limited range and endurance. The small Viper could only be used to fly back to base, as it did not even provide enough power for the SR.53 to take off without additional rocket thrust. Saunders Roe responded with the SR. 177, an improved design in which the 7,300 Newton Viper jet engine was replaced by the more powerful de Havilland Gyron Junior that could deliver a thrust of 32,000 Newton in normal use and almost twice as much with afterburner. This engine was the first British turbojet specifically designed for sustained running at supersonic speeds, enabling the SR. 177 to fly faster than Mach 1 for relatively long periods if assisted by rocket thrust. In addition, a new version of the Spectre engine was to be installed. The Spectre 5 delivered the same maximum thrust as its predecessor in the SR.53 but could be throttled over a wider range down to just 10% of its maximum power, enabling it to be run in combination with the jet engine for prolonged low supersonic flight. The airframe was enlarged to accept the big Gyron Junior, and the positions of the jet and rocket engines were reversed with respect to the SR.53 (i. e. the turbojet was now located below the rocket engine). It also had a larger nose to accommodate the intended new radar system that would assist the pilot to find his target. Like the SR.53, the SR.177 had a pair of Blue Jay (Firestreak) missiles at its wingtips but these could be replaced with pods, each containing 24 unguided air-to-air rockets. An auxiliary power unit with a hydrogen-peroxide steam turbine was installed to power the onboard systems in the event that the turbojet flamed out at high altitude. All in all, the take-off weight of the SR.177 was one-and-a-half times that of the SR.53. Although the general layout was similar to its predecessor (including the 6% thickness ratio of the wings and tailplane) the SR.177 was virtually a new design. The

large Gyron Junior required a lot more air than the Viper, so the sleek lines of the SR.53 with its modest air inlets just behind the cockpit canopy were sacrificed for a huge chin-mounted intake. Of course, the benefits outweighed aesthetics: the new jet engine would considerably improve endurance, enabling the SR. 177 to cruise for a long time on jet power alone. This time could be extended by fitting jet propellant drop tanks under the wings and also by aerial refueling of kerosene from tanker aircraft. The limited supply of rocket propellant could thus be conserved for making a dash once a target was identified. The combined rocket and jet thrust were expected to give the SR.177 a top speed of Mach 2.35 at 12 km (40,000 feet) and Mach 2.75 at 21 km (70,000 feet). For a short-warning interception mission it would be able to achieve Mach 1.6 at 18 km (60,000 feet) in just 4 minutes, but this would consume its entire propellant load and require a gliding return. The powerful jet engine was also expected to enable the SR.177 to be effective in light ground attack and photographic reconnaissance roles, for which the rocket engine would be removed.

In July 1956 funding was secured for the construction of 27 SR.177s, with the first prototype expected to fly by April 1958. It was a large project employing some 400 Saunders Roe and subcontractor engineers on design work alone. Flight testing with the SR.53 was planned to continue, but only to assist in the development of its bigger sister.

However, as elsewhere, changes in military requirements and the evolution of the interceptor concept from a pure rocket plane, to a mixed-power aircraft with a small auxiliary jet engine (the SR.53) and to a mixed-power design with a large jet engine were inevitably leading to the next step of discarding the rocket engine entirely. The death knell for the SR.177 as a Royal Air Force interceptor was the infamous 1957 Defence White Paper in which Minister of Defence, Duncan Sandys, expressed the belief that all manned fighter aircraft would soon be rendered obsolete by guided missiles. Only the development and production of the English Electric Lightning jet fighter would be continued. For the short term this pure jet interceptor was deemed sufficient, because it had become clear that the new Soviet bombers were nowhere near as fast and high-flying as had been feared earlier. No more-capable jet or rocket interceptors would be required, because in the longer term the defense of the country could be safely turned over to surface-to-air missiles. The SR.177 fell victim to what would prove to have been a disastrously erroneous policy that all but killed the UK’s independent military aviation industry. The fact that new, manned fighter aircraft are still being developed and produced today is testimony to how overoptimistic Sandys was about the capabilities of missile systems.

Saunders Roe still tried to sell the SR.177 design to the Royal Navy, as well as to likely foreign customers. The Germans were very interested in its rapid interception capabilities, since their border with the Warsaw Pact meant the Luftwaffe had little time to react to intruding aircraft. When in August 1957 the British decided against the Royal Navy version of the SR.177 the writing was on the wall. Development did continue for a while longer with the British Ministry of Supply agreeing to fund the construction of five of the planned six prototypes in anticipation of sales to Germany. However, the Germans decided not to take financial risks and ordered the

Artistic impression of an SR. 177 in Luftwaffe service.

Lockheed F-104G Starfighter instead, as did many other European NATO countries. This was a pure jet which combined the roles of ground-attack fighter, photo-reconnaissance and interceptor into a single all-weather aircraft. It was later discovered that Lockheed had paid out millions of dollars in ‘Sales Incentives’ (i. e. bribes) to several European pohticians to buy their aircraft and secure what was called the ‘Deal of the Century’. Whatever the reason for the Germans’ change of mind, in hindsight their decision to buy a multipurpose jet fighter instead of a mixed – power interceptor and its derived ground-attack and reconnaissance variants was sensible. The Starfighter was a highly successful, albeit rather dangerous, fighter that remained in operational service with the Luftwaffe until 1987.

With no support left the SR. 177 project was canceled just before Christmas 1957, with the design being 90% complete and the production of various equipment for the first six aircraft already well under way. No complete prototype had yet been made other than a full-scale wooden mockup (unfortunately no longer in existence) which was used for assessing the cockpit layout and the proper fitting of the various pipes, cables and control systems (nowadays this is done by using 3-dimensional computer models, which can be updated and improved much faster than a mockup).

The SR. 53 test flight campaign continued, with the first prototype being joined by the second aircraft with registration number XD 151. This made its first flight on 8 December 1957. Unfortunately it crashed on 5 June 1958 shortly after taking off for its 12th flight. The exact cause was never established but it seems the aircraft lacked the power to climb because its rocket engine had shut down just as the plane left the ground. It struck a pylon at the end of the runway and violently exploded, killing test pilot John Booth. The test program resumed using the first prototype. Saunders Roe proposed turning XD 145 into a true test bed for new rocket engines, including air­launching it (from the back of a Valiant bomber) rather like the experimental X-15 rocket aircraft in the US, but the Ministry of Defence felt the aircraft was not suitable for such a program and halted funding. The decision was not unreasonable, because the SR.53’s aluminum structure would limit its maximum speed to around Mach 2 to avoid structural failure by overheating. Also its air intakes and its still relatively thick wings would probably have caused problems at speeds far over Mach 2. Moreover the aircraft lacked reaction control thrusters for attitude control at extreme altitudes (although these could probably have been retrofitted, as was done to the Starfighter-derived NF-104A described in the next chapter). In October 1958 the flight campaign ended and the only remaining SR.53 was transferred to the Royal Air Force Museum storage facility at Henlow. It remained there until 1978 when it was donated to the Brize Norton Aviation Society, which restored the aircraft for static display. It can now be seen in the Royal Air Force Museum at Cosford. The two SR.53 prototypes together accumulated a total flight time of 22 hours and 20 minutes over 56 flights by three pilots.

During the late 1950s the Spectre rocket engine was also intended to be used in the Italian ‘Leone’ (Lion), which was an attempt by aircraft manufacturer Aerfer to develop their Sagittario 2 supersonic (in a dive) jet into a competitive interceptor, but the project was abandoned before a single prototype could be built.

Saunders Roe had also investigated a third mixed-power interceptor, the rather large P.187. This would have been powered by two Gyron turbojets with afterburners and four Spectre 5 rocket motors, sufficient to enable it to attack targets at Mach 2 within a range of 420 km (260 miles) and up to an altitude of 18 km (60,000 feet). It would have taken 3 minutes and 50 seconds to climb to 12 km (40,000 feet), which was a relatively slow rate of 52 meters per second (171 feet per second). It was to be equipped with an advanced radar and four air-to-air missiles. The propellant required for this performance would have made its take-off weight almost four times that of the SR. 177! The aircraft would have been crewed by a pilot and a radar operator. It featured a sliding nose for improved viewing during landing, an idea that was later adopted by the Concorde supersonic airliner. The P.187 project never left the drawing board.

The cancellation of the SR.53 and SR.177 meant the end of the development of true mixed-power interceptors in the UK but there were also several projects focused on fitting auxiliary rocket engines into existing military jets.

One idea was to put a rocket on the English Electric Lightning jet interceptor, which was still under development when this option was considered. Like on the Mirage III the rocket system would be designed as a self-contained pack that could easily be fitted on the belly of the aircraft if needed. The rocket would considerably increase the maximum altitude of the Lightning, enabling the plane to attack high­flying intruders. The Napier Double Scorpion was developed for this purpose. It had two thrust chambers which burned kerosene in hot oxygen gas that was made by the decomposition of hydrogen peroxide with catalyst packs of metal gauzes and pellets. The engine had a turbopump powered by hydrogen peroxide that was decomposed by a separate, smaller catalyst pack. The total thrust was almost 36,000 Newton, and each combustion chamber could be individually throttled in steps of 0 to 50 to 100% thrust. To test the engine, it was installed on an English Electric Canberra jet bomber. The rocket system, complete with engines, tanks and plumbing, formed an integrated unit that could be carried in the standard bomb bay of the Canberra, thus minimizing the required modifications. To prevent the Canberra from starting to pitch due to the additional rocket thrust from its belly, the nozzles were angled slightly upward so as to aim the thrust through the aircraft’s center of mass. Test flights in early 1956 with Canberra number WK163 started off with a single Scorpion thrust chamber, in which configuration it was even flown at the Paris air show that year. Flights with several versions of the Double Scorpion followed. The system was found to be very reliable and to require relatively little maintenance. A complete overhaul was required after a total firing time of 1 hour (similar to the SEPR 84 series of the Mirage III) but since the Scorpion could only be operated for several minutes per flight, due to the limited propellant supply on board, this was not needed very often. The aircraft successfully demonstrated the Double Scorpion at the Farnborough air shows of 1956, 1957 and 1958, as well as taking part in the 1958 Battle of Britain memorial show.

The Canberra was a subsonic aircraft, and the rocket engine could not be used to increase its maximum speed because the aerodynamics were unsuitable for transonic flight. But it could boost the aircraft’s maximum flight altitude, as it was intended to do for the Lightning interceptor. On 28 August 1957 the rocket enabled WK163 to capture the world altitude record with a flight up to 21.4 km (70,310 feet) under the control of pilot Michael Randrup and observer Walter Shirley. It was the third record attempt: on the first flight the plane had not flown high enough, while on the second the official barograph with which they were to measure the altitude malfunctioned. Experimental rocket planes had flown higher in the US but they were dropped from carrier aircraft rather than taking off on their own power and thus their achievements were not recognized by the FAI as official records. A red scorpion and details of the achievement were painted on WK163’s nose. Now retired after a long career as a flying test bed for all manner of new equipment, this aircraft has been preserved by Classic Aircraft Projects at Coventry Airport and still flies (without rocket power) at air shows in the UK.

Apart from testing engines, there was actually an operational job for which rocket-enhanced Canberra’s proved to be ideal. In the 1950s the British performed atomic bomb tests as part of the development of their own nuclear deterrent. The explosions shot radioactive particles high into the atmosphere and the bomb developers needed samples for study. Because the tests with WK 163 had shown that the rocket engine improved the standard service ceiling of 15 km (49,000 feet) by about 30%, a pair of Royal Air Force B6 Canberras were fitted with Double Scorpions to fly high-altitude missions over Australia and the Pacific and collect samples from test explosions.

One of these aircraft, registration number WT207, is loaned to Napiers to conduct further flight testing of their Double Scorpion propulsion unit. On 9 April 1958 pilot John de Salis and navigator Patrick Lowe are flying a test, climbing from 13.7 km to 18.3 km (45,000 to 60,000 feet) in just under 4 minutes by combined jet and rocket

power (a very modest climb rate, but the mixed-power Canberra was meant simply to reach high altitudes, not rapid rates of climb and high speeds). After shutting down the Double Scorpion, the aircraft descends to 17 km (56,000 feet). At that point de Salis ignites the rocket engine once again in order to burn off the residual hydrogen peroxide to ensure that none of this dangerous Uquid remains in the tank for landing. However, upon ignition the crew immediately notices a rumbling noise and buffeting of the aircraft, and a warning light for the rear bomb bay flips on: the rocket engine is on fire! Before they can start the carbon dioxide fire extinguisher in the bomb bay an explosion blows apart the central section of the plane, ripping both wings and the tail off. Mist appears in the crew cabin due to rapid depressurization. Lowe jettisons the hatch above his seat and ejects first, then de Salis shoots out through the clear canopy above the cockpit. They set a record for the highest ejection seat escape ever, which is only broken in 1966 when the crew of an SR-71 Blackbird escapes at an altitude of 24 km (80,000 feet). The likely cause of the loss of WT207 was a leak in the engine which allowed kerosene and HTP (hydrogen peroxide) to mix and ignite outside the combustion chamber. After modifications to the Scorpion the flight testing resumed, but not for long because the British decided that there was after all little need for an auxiliary rocket pack for the Lightning interceptor (which was already demonstrating a rather phenomenal rate of climb and speed for its day). The Scorpion engine was canceled, and with it all rocket development work at Napier ceased. It also meant the end of the Canberra flights using that company’s rocket engine after a total of 500 rocket propelled flights.

Another engine that was mated with the Canberra was the 36,000 Newton thrust de Havilland Spectre, which also used hydrogen peroxide and kerosene. It was fitted in a Canberra and first ignited in the air in January 1957. The test flights showed that the propellant delivery system required further development work. The Spectre was removed from the aircraft and subsequently fitted on another Canberra. This aircraft was never used for nuclear sampling missions but the test flights cleared the engine for use on the aforementioned SR.53 mixed-power interceptor.

In the mid-1960s the South African Air Force was interested in purchasing the subsonic Blackburn Buccaneer, which was in service in the British Royal Navy as an attack aircraft. The Buccaneer was meant to fly very low in order to avoid detection by radar. However, due to their ‘hot and high’ conditions, the South Africans wanted the aircraft to be equipped with rocket engines for extra thrust during take­off. This resulted in the Buccaneer S.50 version with two retractable Bristol Siddeley BS.605 rocket engines fitted just behind the bomb bay. Together these could give 30 seconds of thrust at 36,000 Newton to supplement the 180,000 Newton from the twin jet engines running on afterburner. As with the Super Sprite and Scorpion, the BS.605 used kerosene and hydrogen peroxide as propellants. The modifications required to the standard Buccaneer design were modest. South Africa bought 16 of the planes, which flew with 24 Squadron from 1965 until 1991. Ironically, the rockets turned out not to be necessary and were eventually removed from all operational Buccaneers. The roar of the twin BS.605 rocket engines was only unleashed four times, during air show displays.

The number of mixed-power interceptors developed in the 1950s and early 1960s is astonishing, especially considering that the results in terms of operationally useful aircraft were fairly disappointing. In those years, Cold War tensions produced large aeronautical development budgets. Remaining at the forefront of technology meant a parallel investigation of every possible means of gaining military advantage. In the end the mixed-propulsion interceptor proved to be a dead end, since improvements in turbojet technology, advanced radars for early detection of airspace intruders, and the growing military desire for multi-purpose fighter aircraft, combined to render placing rocket engines on aircraft redundant. The added performance simply was not worth the disadvantages of shorter range and endurance, greater aircraft weight, and added complexity. And the role of the interceptor itself became much less important during the 1950s as it became possible to use surface-to-air missiles to shoot down high-flying bombers. Moreover, the Soviet Union was developing ballistic missiles to deliver nuclear warheads, and these could not be intercepted by any manned aircraft. Fast, low-flying bombers carrying nuclear weapons remained a serious threat, but rocket propelled interceptors were ill-equipped to do anything about these.

There was still a role for high-altitude, high-speed reconnaissance aircraft (‘spy planes’) but they required endurance to be able to penetrate far into enemy territory and cover an adequate area. Advanced jet aircraft, specifically the Lockheed SR-71 Blackbird, soon filled that role. The SR-71 could not fly quite as fast and as high as some of the latest experimental rocket planes, but it could undertake missions lasting hours rather than minutes.

The mixed-power high altitude aircraft did come back one more time in 1963 in the form of the NF-104A AeroSpace Trainer, a heavily modified F-104A Starfighter with a rocket engine on its tail that was developed to train pilots in flying at extreme altitudes (of which more later).

Had mixed-powered interceptors remained of interest to the military, then their evolution would probably have eventually led to advanced systems combining rocket and jet functionalities into a single engine. Such engines, which operate as turbojets at low speeds and altitudes, ramjets at high speeds, and rockets at high altitudes, are currently seen as the key to making fully reusable single-stage spaceplanes possible. But the demise of the mixed-power interceptor sent jet engines and rocket engines their separate ways, with jet engines powering aircraft and rocket engines powering expendable missiles, space launchers and a few specialized high-speed, high-altitude experimental aircraft.