Category Rocketing into the Future


Just like its axis partner, Japan was seeking a ‘wonder weapon’ that could counter the overwhelming power of the AlUed forces. When Japanese military attaches first saw the Me 163 on a visit to Peenemiinde West in 1943 they were mightily impressed and recognized in the revolutionary aircraft a means of halting the expected onslaught of American bombers. The attaches had seen the devastation bombing had caused in Germany and knew that soon Japan would be facing a similar fate. The high-flying Boeing B-29 Superfortress bomber, already in production, would surely soon darken the skies over Tokyo. Still struggling to develop turbo­superchargers to enable their conventional propeller airplanes to reach the altitude at which the B-29 would cruise, Japan was desperate to find an alternative defense measure (turbo-superchargers compress the thin air at high altitude, so that sufficient oxygen gets into an aircraft’s piston engine for proper combustion).

Commander Eiichi Iwaya (who went on to play a crucial role in this story) saw a Komet demonstration at the Rechlin airfield in April 1944, then wrote in his diary: “The roar, and the blue-green flame from the Walter motor, with its immense 1,500 kg of thrust, was evidence that German technology was still ahve here, even if the combat situation was steadily getting worse.”

In early 1943 the Japanese had negotiated licences to manufacture the Me 163B and its engine, although this did not come cheap: the rights and information to build the HWK 109-509A rocket motor alone cost them 20 million Reichsmark, which is equivalent to something like $100 million today. The Japanese were allowed to study the aircraft’s production in Germany, as well as the operational procedures of the Luftwaffe. Documentation on the Komet was sent to Japan by the German submarine U-511, which left Lorient in occupied France on 10 May 1943. It was then put into service with the Japanese navy under the name RO-500. To enable Japan to build its own forms of the aircraft, Germany later agrees to supply complete blueprints for the Me 163B and the HWK 109-509A, as well as a complete Komet, two sets of sub­assemblies and components, and three complete rocket engines.

This equipment and documentation is loaded onto two Japanese submarines, one of which leaves the German harbor of Kiel on 30 March 1944 and the other leaves from Lorient just over a fortnight later. The first, RO-501, is sunk in the mid­Atlantic by the destroyer USS Francis M. Robinson. The second boat, the 1-29, reaches the harbor at Singapore. Soon thereafter, Allied intelligence intercepts a message from Berlin to Tokyo that lists all the strategic cargo carried by the 1-29. Before leaving Singapore the 1-29 commits the mistake of radioing a detailed itinerary for the final part of its journey to Japan; the message is intercepted and decoded by US Navy’s Fleet Radio Unit. The Navy sends three submarines to hunt the 1-29 and on 26 July the USS Sawfish finds the enemy boat running on the surface near the Philippines. The Navy submarine launches four torpedoes: three hit and sink the 1-29. With all the Me 163 parts and plans on the ocean floor, the Japanese would not have received any of the material sent from Germany in 1944 were it not for a Japanese naval mission member who left the 1-29 in Singapore soon after it docked. Unaware that the 1-29 is doomed by sloppy communications security, Commander Iwaya flies to Japan with a briefcase containing only twenty pages of the design manual, a photo of the Me 163B and another of its wing, a document describing the production and handling of the dangerous propellants, and data for several types of valves used on the plane. It is by no means sufficient to build a copy of the Komet and its advanced rocket engine, but this is all that reaches Tokyo.

For the Japanese aeronautical engineers the scarce information, together with what they received from the U-511 the previous year and radio telegraph communications with the air ministry in Berlin, actually turns out to be sufficient to develop their own forms of the Komet. The project is set up as a joint effort by the Imperial Japanese Army Air Service and the Navy Air Service, but they argue about how to proceed. The Army proposes to develop a new, completely Japanese rocket interceptor that can carry more propellant than the Me 163B in order to increase the duration of its powered flight. The Navy seems to better understand the urgency of Japan’s gloomy military situation, and proposes to stick as closely as possible to the already proven design of the German Me 163B. The Navy Air Service wins the dispute, and in July 1944 specifications are issued. Aircraft manufacturer Mitsubishi Jukogyo KK wins the contract to develop and produce two models of the new plane: the J8M1 ‘Syusui’ (Autumn Water) for the Navy and the almost identical Ki-200 for the Army.

However, the Army secretly decides to develop its own concept, independently of Mitsubishi, at its aero technical institute Rikugun Kokugijitsu Kenkyujo at Tachikawa (as in Germany there is considerable rivalry between the different military services, even although by now the desperate situation of the war leaves little time for such a divergence of effort). This Ki-202 is to be a better plane than the Me 163B, the J8M1 and the Ki-200, but it never leaves the drawing board. It was to have had a stretched fuselage to carry more propellant and used a KR20/Toku-Ro 3 dual­chamber rocket engine similar to the HWK 109-509C, with the large chamber and nozzle producing 20,000 Newton of thrust for take-off and climb, and the smaller one 4,000 Newton for cruise flight.

Work on the J8M1 and Ki-200 rocket plane versions quickly gets underway with Mitsubishi manufacturing the Japanese form of the HWK 109-509A engine, which is known variously as the KR10 and the Toku-Ro 2, and joining forces with Nissan and Fuji to develop the airframe. The Naval Air Technical Arsenal in Yokosuka develops the MXY8 glider ‘Akigusa’ (Autumn Grass), which will initially be used to study the handling characteristics of the Syusui and later to train its pilots. On 8 December, Lieutenant-Commander Toyohiko Inuzuka makes the first flight in an MXY8 glider after being towed into the air from Hyakurigahara airfield, and finds that it closely matches the described handling characteristics of the German Komet. Two additional gliders are made in Yokosuka, one of which is delivered to the Army for evaluation at its own aerotechnical institute.

The resulting J8M1 rocket plane looks very similar to the Komet (as intended) but it has a take-off weight some 430 kg (940 pounds) lighter due to a reduced propellant volume, less ammunition, and the deletion of the armored cockpit glass (the lack of armor protection for pilots and engines was a common feature of Japanese fighter aircraft, resulting in a weight reduction and hence increased agility at the cost of a higher vulnerability to enemy bullets). However, the weight reduction does not fully compensate for the fact that the KR10 engine produces less thrust than the original Walter engine: 14,700 instead of 17,000 Newton. The Syusui will therefore not be as fast as the Komet and will have a slower rate of climb. Nevertheless, its top speed of 900 km per hour (560 miles per hour) will still be more than sufficient to outrun any Allied propeller fighter aircraft, and its rate of climb of 48 meters per second (156 feet per second) is still phenomenal. The German Komet could achieve 960 km per hour (600 miles per hour) and had a rate of climb of 61 meters per second (199 feet per second). The distinctive power-generation propeller of the Komet is omitted from the J8M1, which instead has a longer nose section housing a battery. It is

The Japanese J8M rocket interceptor.

armed with two 30-mm Type 5 cannon that can each spit out 500 rounds per minute. A J8M2 is planned that will differ only in that it sacrifices one cannon for a small increase in the propellant capacity for slightly longer endurance. The Army’s Ki-200 is very similar to the J8M1, the most important difference being that it is equipped with two 20-mm Ho-5 cannon or two 30-mm Ho-155 cannon, each of which can fire as many as 600 rounds per minute.

A production plan is put together that should lead to having at least 3,600 rocket interceptors in operation by March 1946. The first J8M1 prototype is used for load testing on the ground. The next two are for flight testing, and on 8 January 1945 one of these is towed into the air by a Nakajima B6N1 (Allied designation ‘Jill’) bomber from Hyakurigahara airfield. The aim is to test the low-speed aerodynamics, so the plane has no engine or propellant. Water is used as ballast to obtain a realistic total weight and mass balance. The test shows that the design is very good for gliding and should handle well under rocket power at high speeds.

Development of the engine is not going well. The first KR10 prototype explodes immediately upon being started, as does a modified engine known as the KR12 (the KR12 design does not offer any real advantages over the KR10 and its development is halted). Since the Japanese engineers have little experience with liquid propellant rocket engines and have only a limited amount of design and production information on the German technology, they are having difficulty designing the equivalent of the Walter engine and especially the small, high-speed turbopump. The resulting delays mean that by mid-1945 the engine is still not available and the J8M1 still cannot be flight tested under power. Time is running out because the Allied forces are virtually banging on the door of the Japanese home islands. Captain Shibata, commander of the Navy air group which is to be the first to operate the J8M1, tries to speed up the development by agreeing with the development team that the engine will be deemed ready for flight if it can operate for at least 2 minutes without failures; desperation is clearly becoming a driving factor in the plane’s development.

In the meantime, another submarine is dispatched from Germany to deliver Komet documentation and equipment to Japan. Also on board are Messerschmitt engineers Rolf von Chlingensperg and Riclef Schomerus. The U-864 leaves Bergen in Norway on 5 February 1945 but a couple of days out is obliged to return due to trouble with one of its two diesel engines. On the way back her periscope is spotted by a British submarine, HMS Venturer, which has been dispatched to intercept the U-864 and its cargo in response to the interception of German radio transmissions. On 9 February the submerged Venturer fires four torpedoes in a spread pattern at the U-864, which crash dives but suffers a hit by one torpedo and breaks in two. With it the last load of Komet equipment sent to Japan disappears into the depths.

The date for the first Syusui powered flight slips further when another engine prototype blows up. In addition, relocating the KR10 and Syusui development teams lest they be bombed by B-29s results in even more delays. Only in June 1945 does the engine meet the 2-minute thrust requirement. In fact the KR10 development team working in the Yamakita factory runs an engine for 4 minutes. The Mitsubishi J8M1 group at the Matsumoto research center runs another engine for 3 minutes. The striped pattern in the exhaust flame caused by the shock waves in the supersonic gas flow leads the engineers to dub the thrust plume the “tail of the tiger”. Unpowered glide tests of one of the Syusui prototypes with an engine installed, and tests running the engine within the plane’s fuselage, are quickly organized and successfully completed. On 7 July the first J8M1 with an engine of doubtful trustworthiness finally stands ready at Yokoku airfield for its first flight under rocket power. Following the procedure established for the Komet, Lieutenant Commander Toyohiko Inuzuka rolls 320 meters (1,050 feet) down the runway and takes to the air after only 11 seconds, successfully jettisoning the dolly. He then flies horizontally to gain speed prior to climbing at a 45 degree angle. All is well up to this point, but at an altitude of some 350 meters (1,150 feet) the engine sputters, puffs black smoke and quits. The speed of the plane carries it up a further 150 meters (500 feet) whereupon Inuzuka levels off, dumps the remaining propellant and banks to the right intending to glide back to the airfield. However, the maneuver makes the plane lose speed at an alarming rate and causes it to drift off in the direction of a small building. Inuzuka pulls the nose up in a desperate attempt to avoid it but nevertheless clips the side of the structure. The aircraft comes down in a terrible crash. Inuzuka is extracted from the wreck severely injured, and dies shortly after.

It was soon realized why the engine had cut out. For the test flight the propellant tanks had been only half-filled for safety reasons and when Inuzuka started his steep climb the liquids sloshed to the back of the tanks. But the feeds to the engine were at the front of the tanks. This deprived the engine of propellant. The reason for the feed points being at the front of the tanks was that during actual combat operations it was expected that the Syusui would climb above the enemy bombers and then attack in a powered dive. Such propellant as was left in the tanks would then slosh forward, and this had prompted the designers to put the feed lines there. The incompatibility of this system with a powered climb instead of a dive with half-filled tanks had not been noticed before the test flight. It is decided that the propellant supply system from the tanks to the KR10 must be modified, and in the meantime all powered

flight testing is stopped. While the crash investigations are underway, two new KRIOs blow up on the test stand indicating that the engine still has other issues in need of resolution.

However, less than a month after the disastrous test flight the first atomic bomb is dropped on Hiroshima. Another incinerates Nagasaki several days later. In response, Japan surrenders unconditionally.

Component production for the new airplanes had already begun in preparation for mass production of J8Mls and Ki-200s. After the three prototypes, four machines had been produced (one of them the first Ki-200). Training courses for Army and Navy pilots were also being organized using the Ku-53, an engineless glider with the same configuration and flight characteristics as the actual plane. However, even more than in Germany, the revolutionary rocket interceptor came too late to make a difference and was never used in combat.

Only two examples of the J8M1 survive today. One is on display at the Planes of Fame Museum in Chino, California. It is one of two aircraft taken to the US aboard the USS Barnes in November 1945 and which, after evaluations, were sold for scrap. One was apparently turned into pots and pans but 19-year-old Ed Maloney found the other one in a southern California storage yard. The owner of the facility thought it was some kind of boat but Maloney recognized it for what it was and bought it for the cost of its unpaid storage charges. It became the first plane in the museum that he was planning and is now the Planes of Fame Museum. Another Syusui fuselage was discovered in 1961 in a cave in the Yokosuka area south of Tokyo, badly damaged and incomplete. Until 1999 it was on display at a Japanese Air Force Base near Gifu, then it was restored and completed with replica parts by Mitsubishi, whose archives contain 80% of the original blueprints (additional information for the restoration was obtained by studying the Planes of Fame Museum example). This plane can now be viewed at the company’s Komaki Plant Museum, bearing its original bright yellow-green paint.

Towards the end of the war the Japanese Navy were also working on a completely home-grown rocket attack aircraft called the Mizuno ‘Shinryu’ (Divine Dragon). The project started as a fairly conventional-looking design for a simple kamikaze glider that could be launched from the shore using a trio of 1,300 Newton Toku-Ro Type 1 solid propellant rockets with a 10 second burn time. This plane was to crash into and thereby destroy Allied ships, or even tanks if these managed to get onto the beaches of Japan. A prototype for the glider was tested without the rocket engines, and these flights showed that the rocket propelled version could be expected to be difficult to fly. But experienced and well-trained pilots were too valuable to sacrifice in suicide attacks. A revised design was therefore proposed for a nimble attack aircraft powered by four 1,500 Newton thrust Toku-Ro Type 2 solid propellant rockets burning for 30 seconds. Armed with eight unguided air-to-ground rockets it would be able to attack ships and tanks, or even intercept B-29 bombers at a top speed of 300 km per hour (190 miles per hour) without the need for a suicidal collision. The Mizuno Shinryu was a canard design with large swept wings, small vertical stabilizer wings on the nose, and a vertical tail fin; the rocket boosters were to be fitted inside the tail part of the fuselage.

As explained earlier, the vertical stabilizers behind the wings of a conventional airplane push the tail down and compensate for the tendency of the main wings to push the nose down. An obvious disadvantage of this balancing technique is that the vertical stabilizers effectively provide negative lift, which has to be compensated by additional lift from the main wings. On a classical canard design the stabilizers are put in front of the wings, creating balance by pushing the nose up rather than the tail down; both the wings and the horizontal stabilizers provide positive lift, resulting an aerodynamically more optimal design. However if the plane pitches up, the angle of attack on the canard stabilizers increases and this gives them more lift, which in turn makes the plane’s nose rise even faster. A canard airplane can thus be rather unstable in pitch. In the hands of a skilled pilot this translates into high maneuverability but for an inexperienced pilot tends to produce a crash. There are several other pros and cons regarding canard airplanes and different effects depending upon the location of the center of gravity relative to the center of pressure (lift), but it is interesting that a number of highly maneuverable modern jet fighters have shapes remarkably similar to the Shinryu; notably the Eurofighter Typhoon, the Sukhoi Su-30 and the Dassault Rafale.

To enable it to attack B-29 bombers flying at high altitude a pressurized cockpit was proposed for the Shinryu, but a pressure suit for the pilot in combination with an unpressurised cockpit was also possible. The plane would take off with the help of a two-wheeled jettisonable dolly, and it would land using fixed skids under the wings and nose. It seems that other ways of getting the plane into the air were considered, likely involving towing or carrying by other aircraft in order to extend the range and preserve the rocket motors for the actual attack (as was proposed for several German rocket interceptors). Although some sources claim the vehicle was also designed to be used for kamikaze attacks with a warhead fitted into the nose, the complexity of the aircraft, its innovative aerodynamics for high maneuverability, and the presence of landing skids make this improbable. The plane never even reached the prototype stage before Japan surrendered, and even the earlier glider design never flew under power due to problems in developing the required rocket motors.

The only Japanese rocket airplane that actually did see action near the end of the war was the Yokosuka MXY-7 Type 11, which was more of a manned suicide anti­ship missile than a plane. It had a torpedo-shaped fuselage fitted with small, straight wooden wings and a horizontal wooden stabilizer that had a vertical fin at each end; there was also one experimental version built, the Type 21, that had thin steel wings manufactured by Nakajima instead of the standard wooden wings. The cockpit was positioned behind the wings and just behind the 1,200 kg (2,650 pound) bomb which occupied most of the fuselage. The pilot was protected by armor plating beneath as well as behind his seat because it was expected that the plane would encounter anti­aircraft fire from ships as well as from interceptor fighters. Instrumentation inside the cockpit provided the bare minimum of information required for the short flight into oblivion. Five fuses where installed to ensure the warhead would go off after impact. At least one of these was expected to be triggered by the impact shock, then detonate the bomb 1.5 seconds later so that the explosion would occur inside the target ship and cause maximum damage.

An Ohka suicide missile found by US soldiers.

The kamikaze piloted bomb was carried partly inside the modified, doorless bomb bay of a Mitsubishi G4M (‘Betty’) bomber, and dropped within striking range of its target (interestingly, the later X-l in the US would start its mission in a very similar manner, also with the pilot entering the rocket plane inside the bomber shortly prior to the drop). The pilot would start by gliding towards his target and then ignite a trio of Type 4 Model 1 Mark 20 solid propellant rocket motors, each delivering a thrust of 2,600 Newton. He could choose to fire the motors sequentially or simultaneously depending on the range and speed required. In a sharp dive firing the rockets together in combination with gravity could smash the aircraft into a ship at about 930 km per hour (580 miles per hour). This tremendous velocity made it virtually impossible for the ship to aim its anti-aircraft guns and shoot the little plane down; something the gunners often succeeded in when attacked by much slower, conventional kamikaze planes.

The Japanese called their kamikaze weapon the ‘Ohka’ (Cherry Blossom) but the Americans referred to it as the ‘Baka’ (Idiot). The plane was conceived by Ensign Mitsuo Ohta, a transport pilot of the Japanese navy who made the first design aided by professor Taichiro Ogawa and students of the Aeronautical Research Institute at the University of Tokyo. Within weeks of contacting the university, Ohta was able to send plans to the Naval Air Technical Arsenal in Yokosuka complete with drawings, wind tunnel model test results and performance estimates. The Navy was sufficiently impressed to tell the Yokosuka engineers of its First Naval Air Technical Bureau in

Ohka dropped by a Betty bomber.

August 1944 to develop the design into an operational machine. The first variant, and the only that was put into service, was the Type 11. As it would not need to take off by itself, land or fly at low speeds, its wings were kept very small to minimize drag and thereby maximize attack speed. In part because of this, the maximum range of the rocket propelled Ohka when dropped from an altitude of 6 to 8 km (20,000 to

27,0 feet) was only 36 km (23 miles). The slow and vulnerable transport bombers laden with their heavy Ohkas were therefore obliged to fly close to the targets (which often included well-defended aircraft carriers, themselves prize targets) before they could release their Ohkas. Many bombers were shot down by defending fighters long before they could get near enough to the enemy fleet to deploy their Ohka.

Solid propellant rocket motors are simple, cheap and expendable, and therefore a natural choice for a single-mission kamikaze plane. However, the disadvantage of the short range led to several Ohka proposals powered by jet engines for longer flights. None of these alternative designs could be put into operation before the war ended.

Unmanned tests of unpowered and powered prototypes were followed by the first manned Ohka flight on 31 October 1944. On that day Lieutenant Kazutoshi Nagano straps into the prototype of the Ohka K-l trainer, a version that had water tanks in the nose and tail instead of a bomb and rocket motors. For this test it has been equipped with two small solid propellant rocket boosters, one under each wing. Nagano is dropped by a Betty carrier plane at an altitude of 3.5 km (11,500 feet), pursues a stable glide for a few minutes and then fires the two motors. Immediately the machine begins to yaw, so Nagano quickly jettisons both rockets and manages to find a stable glide position once more. It is later found that the two rockets had not yielded the same amount of thrust and their positions on the wings caused the stronger rocket to attempt to turn the plane around its vertical axis (a similar problem had plagued the German Natter). Shortly prior to landing Nagano drains the water tanks in order to reduce the plane’s weight and make it possible to fly and land at a relatively low speed on skids fitted beneath the fuselage and the wings (the operational Ohkas would of course not have an undercarriage, since they were not supposed to come back). Other than during its brief moment under rocket power the airplane handled well. It is decided not to use wing-mounted rocket motors anymore, and only equip the Type 11 with three in the tail, very close to the centerline so that any difference in thrust between the boosters will have little effect on the control of the vehicle. Subsequent test flights established how the Ohka should be operated. After being released from the bomber, the Ohka pilot would enter a shallow glide at a speed between 370 and 450 km per hour (230 and 280 miles per hour). Between 8 and 12 km (5 and 7 miles) from the target, and at an altitude of about 3.5 km (11,500 feet), he would ignite all three rocket motors and accelerate to a speed of 650 km per hour (400 miles per hour), then put his machine into a 50 degree dive to scream down at 930 km per hour (580 miles per hour), at the last moment pulling up the nose in order to hit the targeted ship at the waterline.

A navy flight unit was set up to operate the Ohka. Soon nicknamed the ‘Thunder God Corps’ it drew hundreds of volunteering pilots in spite of the nonexistent career prospects. After rejecting those who were married, were only sons, had too important family responsibilities, or were simply too old, some 600 remained. They trained first with a conventional Mitsubishi A6M ‘Zero’ fighter, practicing the attack profile with the engine off. Several pilots were given some flight instruction in one of the MXY7 prototypes of the K-l two-seat trainer, but most were only able to rehearse using an Ohka while it sat on the ground.

It appears that some 751 Ohka Type 11 aircraft were built at two production sites but only a small number were deployed against the Allied fleet before the war ended, with poor results. The first time they are used is on 21 March 1945. Sixteen Betty bombers carrying Ohkas and two Bettys to provide navigation and observation are sent to attack US Navy Task Group 58, which includes the aircraft carriers Hornet, Bennington, Wasp and Belleau Wood. The bombers were to have been escorted by a force of 55 Zero fighters but owing to technical problems 25 of these either did not take off or had to turn back. Some 113 km (70 miles) from their target the planes are intercepted by 16 F6F Hellcat fighters. The Bettys immediately jettison their Ohkas, sending the pilots to watery graves without even a chance to impart damage on the enemy. All the Bettys are shot down and only 15 of the escorting Zeros make it back to base. On 1 April six Betty/Ohka combinations try again, attacking the US fleet off Okinawa. This time the bombers are able to approach close enough to their targets, and one Ohka completes a successful attack on the battleship West Virginia, causing moderate damage to one of its gun turrets. Three cargo ships are also hit by suicide aircraft but these may have been conventional kamikaze planes rather than Ohkas. The Bettys are all shot down. Eleven days later nine Bettys with Ohkas again attack the US fleet off Okinawa. One Ohka plunges onto the destroyer Mannert L. Abele, causing an explosion that rips the ship apart. The Abele becomes the first warship to be sunk by a kamikaze rocket plane, demonstrating the vulnerability of a ship to an accurately guided rocket missile flying at high speed. Another Ohka aiming for USS Jeffers is hit by anti-aircraft fire from the destroyer and blows up only 45 meters (148 feet) from the ship, imparting extensive damage. A pair of Ohkas target the destroyer Stanly and one hits it just above the waterline. However, the plane’s explosive charge punches right through the ship and explodes only after emerging from the other side of the hull, causing little damage. The other Ohka narrowly misses the ship and falls into the sea. Only one Betty survives the mission.

Further attacks on 14 April (with seven Ohkas), 16 April (six Ohkas) and 28 April (a night attack with four Ohkas) fail to produce any Ohka hits and most of the Betty carriers are shot down. On 4 May the Japanese have better luck when one of seven Ohkas sent against the US fleet near Okinawa hits the bridge of the destroyer Shea and causes extensive damage and casualties. The minesweeper Gayety is damaged by a near-miss of another Ohka. Again the price for the Japanese is high, as all but one Betty is lost (in addition to all Ohkas and their pilots). On 11 May one of the four Ohkas sent out hits the destroyer Hugh W. Hadley and causes such extensive damage and flooding that the vessel is deemed to be beyond repair. But this proves to be the final Ohka success, as the attacks on 25 May (eleven Ohkas, with most returning to base owing to bad weather) and 22 June (with six Ohkas, two of the Bettys making it home) are ineffective. Initially presuming the Ohka to be just a new type of anti-ship bomb, the Allies learn of its true nature only after capturing some of the machines on Okinawa in June 1945.

Deployed in greater numbers, the Ohka might have played a more significant role in the war but the few successes do little to stop the mighty US fleet. These attacks do show how difficult it is to stop a rocket propelled missile, and that a single hit can severely damage or even sink a large warship. The post-war development of anti-ship missiles was rapid but the kamikaze pilots were superseded by electronic guidance and the rockets were launched from fast attack jets, not lumbering bombers like the Betty whose crews had also been effectively flying suicide missions.

In an artificial cave set behind the main buildings of Kenchoji Zen Temple in the ancient Japanese capital of Kamakure there is a monument to remember the Ohka missions. A steel plaque lists all of the Ohka pilots as well as the crews of the Betty bombers who died in the first ever (and hopefully final) kamikaze attacks using a rocket plane. Another plaque in the cave tells the Ohka story, albeit from a rather nationalistic Japanese perspective. The final part of the engraved text reads: “.. .Ohka attacks together with special attacks by Zero fighters carrying bombs were made repeatedly. That heroic battle tactic made American officers and men tremble

with fear. This monument to the Jinrai warriors honors those pure and excellent young men who, without regard for their own sacrifice, courageously went to their place of death for their homeland and fellow countrymen.”

Many Ohka Type 11 have survived and are on display at museums in the US, UK and Japan. The only remaining Type 22 jet-propelled version is at the National Air and Space Museum’s Steven F. Udvar-Hazy Center, which also has a dual-cockpit trainer that seems to have been intended for preparing pilots for land-based launches, where Ohka’s would have taken off with the help of a rail launch cart equipped with two solid propellant rocket boosters.

Japan also experimented with rocket boosters to assist aircraft take off. They were applied to the Nakajima ‘Kitsuka’, a Japanese version of the Me 262 jet fighter that suffered from underperforming jet engines. It was hoped that the additional power of jettisonable rocket boosters would limit the otherwise very long take-off run of this plane. On the second test flight of the first prototype, four solid rocket boosters with a thrust of 8,000 Newton each were installed. However, they had not been set at the correct angle. Rather than adjust them, which would take too much time, the thrust of each booster was halved in the expectation that at such low thrust the misalignments would not cause a significant disturbance to the balance of the plane. On 11 August 1945, after a delay of one day owing to the high level of activity of enemy aircraft in the vicinity, pilot Lieutenant-Commander Susumu Takaoka climbs into the cockpit, has the engines started and then taxies onto the runway where he stops, extends the wing flaps for additional lift at take-off and opens the throttles of the jet engines to build up thrust prior to releasing the brakes. Four seconds into the take-off run he ignites the four boosters and promptly finds himself in serious trouble. The sudden thrust of the four downrated boosters forces the nose of the plane up and makes the tail slam down onto the runway. He pushes the stick forward but it doesn’t help. The boosters burn for 9 seconds with the plane essentially out of control. Just one second before the units bum out, the plane’s elevators suddenly take effect and slam the nose down hard. Takaoka decides to abort the flight, but his brakes have little effect. He mns off the mnway, the undercarriage collapses upon encountering a drainage ditch, and the plane slides to a standstill on its belly. The damage is severe, not only to the landing gear but more importantly to the engines slung under the wings. It is likely that the misalignment of the boosters was the cause of the problems. The Kitsuka never flew again because Japan surrendered several days after this aborted test flight and work on the project halted.

Rocket boosters were also envisaged to help the egg-shaped Kayaba ‘Katsuodori’ get up to take-off speed. This was to be a fighter plane powered by a ramjet engine. As explained earlier, a ramjet is a jet engine without a rotating compressor or turbine in which the air is instead compressed by the speed of the airplane and the shape of the air intake (in other words, the air is rammed into the engine as the plane flies through the atmosphere, increasing its pressure sufficiently for proper combustion and effective thrust). Owing to the lack of moving parts the design of a ramjet is relatively simple, is more reliable than a gas turbine engine, and ought to require less maintenance. By drawing oxygen from the air it can ran for much longer than a rocket engine on the same amount of propellant. But for the ram effect to compress the air sufficiently for the engine to start to function, the plane must already be flying at a speed of at least Mach 0.3. Hence the engine could not be used for take-off. To get the Katsuodori up to speed, four solid propellant rocket boosters were to be mounted on the side of the fuselage beneath the wing roots, in pairs, one above the other. During the horizontal take-off run on an ejectable wheeled dolly the rockets would be fired in pairs, one on each side of the plane, with each lasting 5 seconds. After the second pair burned out the plane’s speed would be sufficient for the ramjet to operate and all of the boosters would be jettisoned. But the project was shelved in 1943, probably in part due to the danger of using multiple solid propellant boosters: they cannot be stopped and if one misfires or has a lower thrust than its partners then the plane can quickly lurch out of control (recall that the German Luftwaffe used liquid propellant Walter rocket pods rather than solid propellant boosters to help their heavy bombers take off).

As in Germany, near the end of the war Japan also had a simple rocket aircraft for ramming purposes under development. Like the Zeppelin ‘Rammjager’, its pilot was to ram his rocket powered plane into an enemy bomber, then glide back for a landing (followed by another hair-raising ramming mission in the same machine until either it or he were lost). But unhke the German machine it was not armed. It would have been accelerated by four of the same solid propellant boosters as used by the Ohka. While the Ohka had only three motors and was laden with a massive warhead, four rockets with a combined thrust of 11,000 Newton would probably have been able to drive the light ramming plane to a speed of over Mach 0.9. The concept remained a paper study but if this plane had taken to the air its test pilot would likely have found himself in serious trouble since the Japanese had virtually no knowledge of transonic aerodynamics. The plane had swept wings but it is not clear whether the designers understood the benefits of such wings at transonic speeds or whether they introduced this shape merely to provide an angled edge to cut more easily through the tail of an enemy plane. It is not known whether the Japanese rammer was to be towed into the air like its German counterpart or be launched from the ground.

The US had no appreciation of how far Japanese jet and rocket plane technology had advanced until after the surrender, when they got access to the aircraft factories, many of which were hidden in tunnels in the mountains, and discovered the fleet of advanced airplanes that Japan had been busy building in preparation for the defense of their home islands. Had Japan started its advanced projects a bit earlier, and had the US been forced to invade Japan itself by conventional military means rather than forcing a surrender by dropping atomic bombs, the US ships and planes would have been met by a variety of jet and rocket planes against which they would have had little defense.


The basic, modern rocket spaceplane concept was bom in the 1930s. It involved a hypersonic rocket aircraft launched by a rail with the assistance of a rocket sled, and capable of flying almost around the globe by using a suborbital trajectory in which it repeatedly bounced off the upper layers of the atmosphere, like a flat stone skipping over the surface of a pond. This ‘rocket glider’ idea was way ahead of its time, and in fact is still ahead of our time. The ‘SilbervogeP (German for ‘Silverbird’) was the brainchild of Austrian-German aerospace engineer Eugen Sanger. He first published

M. van Pelt, Rocketing into the Future: The History and Technology of Rocket Planes, Springer Praxis Books, DOI 10.1007/978-1-4614-3200-5 8, © Springer Science+Business Media New York 2012

the idea in his 1933 book Raketenflugtechnik (Rocketflight Technology). In fact, this work was an elaborated version of his original plan for a doctoral dissertation, but his university had advised him to find a more classical topic and he subsequently got his degree for a study of aircraft wings. Rocket pioneer Hermann Oberth had shown that a rocket plane could only hope to achieve a long range by starting with a very steep ascent, leveling off at high altitude in order to use the remaining propellant to attain its maximum speed in the thin air, and then glide back down for the remainder of its flight. Sanger’s calculations led him to the same conclusions. His initial rocket plane design involved an aircraft shaped like an elongated bullet with large but thin double-wedged wings (like those of the X-2 subsequently built in the US), a small single tail fin and two small horizontal stabilizers. It bore little resemblance to the blunt-shaped biplanes of that time.

Together with Austrian-German mathematician and physicist Irene Bredt, who he would marry in 1951, Sanger further iterated the design during the 1930s and early 1940s. Initially they intended their plane to transport passengers at hypersonic speeds all over the globe but he also tried to pitch the idea as an intercontinental bomber to the Austrian military in 1933. Not surprisingly, in an age where transonic jet airliners were still science fiction the proposal was rejected, primarily because of the risk of the rocket engine exploding.

The German Luftwaffe, however, did see value in a ‘RaketenBomber’ (Rocket Bomber), as well as the basic rocket engine development and testing Sanger was already conducting at the University of Vienna. Not wanting to be outdone by the Wehrmacht (Army) and its rocket development by von Braun’s team, the Luftwaffe in 1936 invited Sanger to establish a secret rocket propulsion research institute in Trauen, Germany. It was named “aircraft test site Trauen” to hide its true nature, and Sanger was even given a fake job at the University of Braunschweig so that people wouldn’t wonder where he was working. Like Wemher von Braun, Sanger funded his work via the military as this was the only source available in Germany during the war. Under his leadership, the Trauen site quickly became an impressive laboratory with a wind tunnel, a rail launch track test rig, and a large rocket engine test stand. Sophisticated experimental rocket engines that ran on fuel oil and liquid oxygen were developed and tested, including components for a 1,000,000 Newton engine intended to propel the Silbervogel. An innovative feature of Sanger’s engine designs was the cooling of the combustion chamber by circulating propellant around it. This cooling strategy, sometimes called the ‘Sanger-Bredt design’, was actually implemented in early rocket engines such as that of the A4/V2, and in most liquid propellant rocket engines since then (including, as we have seen, those used to propel aircraft). For the engine of the Silbervogel, Sanger proposed to use a water loop to cool the massive nozzle and combustion chamber because the propellants he intended to use (fuel oil and liquid oxygen) were poor coolants. The water would be turned into superheated steam by the high temperatures in the engine and then be fed into a turbine to run the propellant pumps and the water coolant pump. After the turbine, the water would be condensed and pumped back into the cooling loop. But having three types of liquids (fuel, oxidizer and coolant) and all the associated pipes, valves and tanks made it a relatively complicated engine.

Although the Nazi government was interested in the Silbervogel as a possible ‘Amerika Bomber’, for which a number of more traditional aircraft designs were also developed, it was clear that the concept was much too advanced for the technologies available in the 1940s and that the bomber would never be ready in time to affect the war. Even Sanger himself predicted that it would take at least 20 years to make his machine operationally useful. The institute at Trauen was closed down in 1942 when the military gave priority to the VI pulsejet bomb and A4/V2 rocket developments at Peenemiinde, because these would be able to be introduced sooner and hence make a greater contribution to the war effort.

Sanger and Bredt transferred to the DFS institute in order to continue their studies and experiments on ramjet technology. But the Silbervogel never attracted sufficient support and funding, and they ended up spending most of the war developing basic rocket engine technology and a design for a ramjet fighter that never got to fly either. They nevertheless always kept thinking about their favorite concept in private and in 1944 summarized all their design and experiment work in a secret report titled ‘Uber einen Raketenantrieb fur Fernbomber’, subsequently translated and published in the US as ‘A Rocket Drive For Long Range Bombers’. The concept they finally arrived at was a 100,000 kg (220,000 pound) rocket plane that would be 28 meters long with a wingspan of 15 meters. It featured a pressurized cockpit for a single pilot near the nose, a central bomb bay, and the main rocket engine as well as two smaller auxiliary engines in the tail. The rest of the fuselage would mostly consist of two parallel rows of tanks containing a total of 90,000 kg (200,000 pounds) of liquid oxygen and fuel oil. Two short wedge-shaped wings with swept-back leading edges and a horizontal tailplane that had a small fin on each tip give the Silbervogel a conventional-looking rocket plane shape. However, the design incorporated the innovative idea of having the fuselage itself generate lift through its ‘flat iron’ shape, hence acting as a partial lifting body.

A powerful rocket sled would be used to help the Silbervogel take off from a long rail track. The sled’s rocket engines would not need to be very efficient, since they would only operate for a brief time and would not leave the track with the plane. For the sled, trading efficiency and thus propellant weight for more brute power in order to get the Silbervogel going would involve little penalty. Sanger proposed to use von Braun’s A4/V2 rocket engines. The starting speed given by the innovative rail launch system resulted in a smaller, lighter spaceplane requiring a less powerful engine and smaller wings in comparison to a similar vehicle that had to take off on its own (at higher speeds the same amount of lift can be provided with smaller wings). The rail system also meant that the Silbervogel would only need a light undercarriage for its glide landing with empty tanks, thereby further reducing its weight. Furthermore, the rail sled would make detailed knowledge of transonic flight behavior unnecessary, as it would push the plane through this mysterious region of aerodynamics whilst it was still firmly connected to the track (as mentioned above, transonic aerodynamics were not properly understood until well after the war). The rail launch system would limit the possible launch azimuths, but Sanger suggested that it could be used with either end as the starting point. As Germany’s enemies were primarily to the west (the US) and the east (the USSR), a single launch track

Illustration of the Silbervogel in a translated version of the original 1944 report.

aligned east-west would probably have been adequate. The spaceplane would be able to alter its flight direction after leaving the track, but the larger the maneuver the more costly it would be in terms of energy and thus range. The aircraft, the rocket sled, the track and the other infrastructure on the ground would all be reusable.

The 1944 Silbervogel design was to use a 6,000,000 Newton thrust rocket sled to accelerate to a speed of 1,800 km per hour (1,100 miles per hour) on a rail track 3 km (2 mile) long. As this would occur in only 11 seconds it would expose the plane and pilot to a substantial but certainly manageable acceleration of 4.6 G. Leaving the first stage sled behind (which would have brakes to rapidly decelerate to avoid flying off the track) the Silbervogel would climb unpowered to an altitude of 1.7 km (5,600 feet) while turning into the intended direction of flight, then continue gliding up at an angle of 30 degrees. At an altitude of 3.7 km (12,000 feet), 25 seconds after take-off, the onboard 1,000,000 Newton rocket engine would fire and in about 8 minutes push the aircraft up to an altitude of 150 km (500,000 feet). Leveling off, the Silbervogel would then continue to accelerate to a speed of 22,000 km per hour (14,000 miles per hour). At the end of its powered run the aircraft would have consumed almost all of its propellant (90% of its take-off weight) and under the constant thrust of its rocket engine reached its peak acceleration of 10 G, which is about the tolerance limit for a trained pilot. In the climbing phase of the flight the G would build up gradually, as gravity would partially counteract the rocket’s thrust, but in level flight the increase from 7 G to 10 G would occur in a mere 20 seconds. Nevertheless, the pilot would sit upright with his back to the rocket engine so that his heart would not have to pump

eled drive plin*

Illustration of the Silbervogel rail launch system in a translated version of the original

1944 report.

blood into his brain against the direction of the acceleration. Knowledge gained from Luftwaffe acceleration experiments on human volunteers and drugged primates led Sanger to believe that such acceleration would not cause a well-trained pilot seated in this position to black out.

The ultimately attained speed would be some 6,000 km per hour (3,800 miles per hour) short of that required to enter orbit, so after reaching its highest altitude the Silbervogel would gradually descend into the stratosphere. However, at an altitude of 40 km (130,000 feet) the increasing air density would generate enough lift to cause the plane to “bounce” back up to about 125 km (410,000 feet) altitude. Part airplane, part satellite, the Silbervogel would repeat this profile a number of times, with each successive bounce getting shallower and covering less distance horizontally owing to the continuous loss of speed resulting from the aerodynamic drag. The aircraft would heat up each time it hit the atmosphere but have time to cool down (by radiating the heat away) during each hop back into space.

The pilot would not have much window space to admire the view because to keep the external shape smooth, and probably also the difficulty of producing windows for extreme temperatures, Sanger foresaw the pilot being provided with side view slits and “optical aids” (presumably periscopes). However, for landing, he noted, “a kind of detachable windshield can be used, since then the pressurization of the cabin and maintenance of the bullet-shape are unimportant”. Exactly what this would have involved is not entirely clear, but it seems to suggest the pilot was to open the cockpit and stick his head out when approaching the airfield; a rather awkward and archaic way to land a hypersonic spaceplane.

Sanger’s baseline hypersonic bomber, equipped with a rocket engine capable of a specific impulse of 300 seconds, would have dropped a rather small bomb of 300 kg (650 pounds) on New York, then continued its unpowered hop-fly over the continent to land in Japanese-held territory in the Pacific. The Silverbird could only lay a small explosive egg, but the psychological and hence political impact of a German bomb of any size dropped on US soil would have been substantial (as was the first US raid on Japan). With a more efficient engine and/or closer targets, the bomb load could have been increased to several metric tons. After landing as a glider (like the Space Shuttle Orbiter) the aircraft would be re-launched back to Germany, dropping another bomb on the way. From point to point, the range would be a spectacular 19,000 to

24,0 km (12,000 to 15,000 miles). Sanger proposed that it would be possible to stretch this distance even further and enable a landing back at the launch site if the efficiency of the rocket engine could be increased. One possibility to achieve this would be to add metal particles to the fuel in order to increase the exhaust temperature, improving the specific impulse to 400 seconds and so either producing higher thrust with the same propellant consumption or running the engines at the same thrust for longer without increasing the tank volume. The basic idea was sound, because metal particles were later added to solid propellants to increase the performance of boosters such as those used by the Space Shuttle. However, for liquid propellant engines it turned out to be impractical since particles can only be kept well mixed with liquid propellants if they are turned into a gel, and doing that leads to all manner of complications.

The feasibility of bouncing off the atmosphere was inadvertently demonstrated in 1962 by Neil Armstrong when flying the X-15. He came down from a high altitude with an overly steep angle of attack (i. e. with the aircraft’s nose too far up), skipped back up and as a consequence severely overshot his planned landing site at Edwards. In fact, he flew all the way to Los Angeles. However, he managed to make a gliding turn with a huge radius and thus fly back to base. Although he bounced only once, he proved the extended range of a Silbervogel-type trajectory. Crewed space capsules returning to Earth can use similar ‘skip re-entry’ trajectories, typically bouncing only once. The Apollo capsules were capable of flying such a ‘skipping’ trajectory, but it was never an operational necessity. However, the Soviet unmanned Zond prototype lunar mission capsules did demonstrate it successfully during several flights.

A postwar analysis of the aerothermodynamics of the Silbervogel discovered that Sanger and Bredt had made an error in the heat load calculations: during the first re­entry into the atmosphere the structure of the spaceplane would have become much hotter than they predicted and consequently would have needed additional protection in the form of heat shield material which would have made the vehicle much heavier than the original design, leaving little room in terms of weight for people or bombs.

After the war Sanger and Bredt were recruited by the French government to work on missiles and ramjets. In 1949 Sanger founded the Federation Astronautique, and in 1951 he was made the first president of the International Astronautical Federation that hosts the annual International Astronautical Congress. Meanwhile in the Soviet Union, Stalin had become fascinated by the Silbervogel concept, which had come to his attention through three captured copies of Sanger and Bredt’s highly secret 1944 report. Stalin saw it as a possible means of attacking the United States because the advent of nuclear weapons meant that the puny bomb load of Sanger’s design could now yield a very powerful punch. In 1947 Stalin sent his son Vasily together with aviation experts Serov and Tokayev to either win over or kidnap Sanger and Bredt and bring them to Russia to further develop their concept. However, Vasily overly enjoyed his playboy life in Paris and Tokayev defected to the British and blew the operation, so nothing came of it. By then, however, the design bureau NII-1 NKAP, established by Mstislav Keldysh in 1946 specifically to work on a Soviet version of the Silbervogel, had further elaborated on the idea presented in the captured report. By 1947 they had concluded that with the rocket engines and structures which either already existed or were likely to become available in the near future, some 95% of the initial weight of the spaceplane would have to be propellant; Sanger and Bredt had calculated ‘only’ 90%. This would leave very little weight for the structure, tanks, engines, and other essential equipment. The poor ratio between take-off weight and structural weight is a well known issue for any spaceplane concept, even today, and particularly so for a pure rocket propelled vehicle.

The Russians decided that ramjets using atmospheric oxygen instead of internal oxidizer would make the design more feasible: the weight limitation of the vehicle without propellant (i. e. the dry weight limit) would increase from 5 to 22%, while the lower maximum speed in comparison to the Silbervogel would still give the plane an intercontinental range of 12,000 km (7,500 miles). This so-called ‘Keldysh Bomber’ would have a take-off weight of about 100,000 kg (220,000 pounds). It would start with an 11 second run along a track that was 3 km (2 mile) long, pushed by a rocket sled powered by five or six liquid oxygen/kerosene RKDS-100 rocket engines with a combined thrust of 6,000,000 Newton. The bomber would climb to an altitude of 20 km (66,000 feet) and accelerate to Mach 3 using a single RKDS-100 and a pair of wingtip-mounted kerosene ramjets. The ramjets would then flame out due to a lack of oxygen and possibly be jettisoned to shed their weight. Next, the rocket engine would push the vehicle to a speed of 18,000 km per hour (11,000 miles per hour) and up into space, after which there would be a series of atmospheric skips just like the Silbervogel. It was estimated that developments in rocket and aircraft technologies would make it possible to start serious development of the design in the mid-1950s. But by then the concept of the intercontinental ballistic missile, which was easier and cheaper to develop and did not require a far-away landing site, had made the Keldysh Bomber obsolete.

Walter Domberger, the former head of A4/V2 development in Germany, tried to interest the US military in the Silbervogel concept when he joined Bell Aircraft in 1950, carefully calling it the ‘Antipodal Bomber’ rather than its wartime moniker of the ‘Amerika Bomber’ (‘antipodal’ meaning two sites exactly opposite each other on the Earth’s surface; more or less the Silbervogel’s maximum flight distance). But the sheer technical complexity of advanced materials, hypersonic aerothermodynamics and guidance accuracy, together with the doubtful military value of such an aircraft, meant that a real project never materialized. Nevertheless Dornberger’s lobbying did eventually lead to the X-15 rocket plane, the X-20 Dyna-Soar concept and ultimately the Space Shuttle. (The Shuttle is not a true rocket plane and it doesn’t use a skip re­entry trajectory but it does resemble Sanger’s design in that it has a flat base, straight fuselage sides, a rounded top, and a series of massively powerful rocket engines in the back.) The X-15 and the Shuttle incorporated several features that Sanger had not foreseen but whose inclusion would have increased both the weight and complexity of the Silbervogel. One example is a reaction control system to control the spaceplane outside the atmosphere and to properly orientate it for re-entry. Another is the need for precise guidance control sensors and electronics. Although one X-15 pilot managed to fly a manual re-entry when his guidance control system failed, for a safe re-entry both the X-15 and the Shuttle required guidance computers that were not available in the early 1950s. Such assistance would certainly have been required for the complicated bouncing trajectory of the Silbervogel.

As a passenger transporter, the original role Sanger intended for his rocket plane, the Silbervogel would have been much too complex and expensive because of its huge propellant consumption, its elaborate take-off and landing infrastructure, and its payload mass and volume allowing only a few passengers. These considerations still plague recent concepts for hypersonic airliners such as the US National Aerospace Plane (NASP) of the late 1980s. The fact that even the less complex and more economical Concorde never reached its financial break-even point makes it difficult to see how a hypersonic passenger transport vehicle could ever be a financial success.

Sanger returned to Germany in 1954 and three years later became director of the newly-created Institute for the Physics of Jet Propulsion in Stuttgart. However, when in 1961 the German government found out that Sanger was secretly making trips to Egypt, presumably to assist a group of rogue German engineers to develop missiles that could be used to attack Israel, they forced him to resign. The loss of his position, and scant prospect that his Silbervogel would be built during his lifetime, sent him into a state of depression and his health rapidly deteriorated. Sanger was unofficially rehabilitated when it was realized he had only acted as a consultant for an Egyptian meteorological sounding rocket project and had lectured at the university of Cairo. He then continued his work by assisting with a new spaceplane project at the revived Junkers company. Based on decades of research, he proposed a 200,000 kg (450,000 pound) reusable, two-stage vehicle consisting of a delta-wing carrier rocket plane and a smaller delta-wing orbital spaceplane. A steam-rocket sled was to provide a starting velocity of about 900 km per hour (560 miles per hour) at the end of a track 3 km (2 mile) long. The steam rocket would be relatively simple, using only heated water, but not very efficient in terms of performance. However, as with the original Silbervogel’s sled, that would not be particularly important for the overall spaceplane system’s performance. Its simplicity would save development effort and cost, whilst making it reusable would be relatively straightforward due to the easy – to-handle and cheap propellant (water) and the absence of extremely hot exhaust gases to eat away at the nozzle.

The orbital vehicle, called ‘HORUS’ (Hypersonic ORbital Upper Stage), would separate from the carrier at an altitude of 60 km (200,000 feet), then proceed into a 300 km (980,000 feet) orbit while the carrier plane returned to Earth. Having fulfilled its mission, HORUS would de-orbit and glide back to Earth very much like the Space Shuttle Orbiter. Both planes would be powered by similar rocket engines, three on the carrier and one on the orbital plane, each engine delivering 2,000,000

Newton of thrust at sea level and burning the powerful combination of liquid oxygen and liquid hydrogen. This is another similarity with the later Space Shuttle, whose three main engines used those propellants and each had a sea-level thrust of

1,800,0 Newton. In fact, the entire concept has similarities to some of the Space Shuttle designs of the early 1970s. Designated the Junkers RT-8, this reusable launcher was to be able to place a 3,000 kg (7,000 pound) payload into low orbit. An improved version would later use a single-stage spaceplane with an integrated rocket-ramjet engine.

In October 1963 Sanger also accepted a professorship at the Technical University of Berlin where in February 1964, aged 58, he died while giving a lecture. Shortly before his death he wrote that he believed the US and the USSR would direct their full technological capacity towards an ‘Aerospace Transporter’ similar to the RT-8 as soon as the ‘Moon Race’ ended and that “there is therefore at the moment a unique but only a short-lived opportunity for Europe, with its great intellectual and material resources, to become active in a sector of spaceflight in which the major space powers have not yet achieved an insuperable lead”. His last spaceplane design, which after his death evolved into the ‘Sanger-Г RT-8-02, would take off vertically without the need for a launch sled and rail, but it never left the drawing board. Junkers was unable to secure enough support from the government since the ambitious, expensive and economically risky project was deemed to be a step too far.

Like Max Valier, but unlike most other rocket and spaceflight enthusiasts of the time, Sanger was a real rocket plane pioneer: he fully intended his rocket engines to propel rocket aircraft rather than ballistic rockets such as the A4/V2. Sanger was certainly a genius but, Uke Wernher von Braun, his willingness to work for the Nazi military in order to move his dream machines closer to reahty, even if it meant they would be used to bomb innocent civilians, has made him a rather controversial figure (his 1944 report contains a map of New York displaying the expected efficiency of a bombing campaign by a fleet of Silbervogel planes). Irene Bredt worked for the institute in Stuttgart until 1962 and survived her husband by 19 years. In 1970 she was honored with the Hermann Oberth Gold Medal for her scientific contributions.

Sanger and Bredt’s ideas have certainly influenced rocket plane design since the end of the war: the idea for the Space Shuttle can be traced back (via Dyna-Soar and the Bell ‘Antipodal Bomber’) to their revolutionary design. In his 1963 book, which was published in 1965 in English as Space Flight; Countdown for the Future, Sanger cites all of the reasons why, in his opinion, “aerospace planes” hold the future of spaceflight instead of expendable missiles: lower costs per flight, the ability to flight- test each type of production vehicle, the possible use of airfields in densely populated countries, easy self-transportation of the vehicle from the production plant to launch site, and the ability to use them for both orbital missions and long-distance flights. These are still the principal reasons for pursuing spaceplanes although nowadays the need for (and the advantages of) such vehicles seem less obvious. Certainly nothing resembling the Silbervogel has yet been built. In the same book he writes, “military aerospace planes will be used as reconnaissance planes, fighter planes against Earth satellites and extraterrestrial space stations, satellite inspections planes etc” and also “we can possibly count space fighter planes among the most important of all space

weapons”. At that time work on the X-20 Dyna-Soar was still ongoing, but after its cancellation work on such military orbital vehicles all but ceased (with the exception of the Space Shuttle, which was partially based on military requirements). However, considering the current renewed interest by the US Air Force in hypersonic vehicles Uke the X-51 and the Falcon Hypersonic Technology Vehicle, as well as automated orbital shuttles like the X-37, Sanger may well have been right on this after all.


The development of rocket planes in Russia before the war followed a very similar path as in Germany, with spaceflight enthusiasts developing rocket motors and later incorporating them into simple gliders. However, whereas in Germany rocket plane development was not deemed to be very urgent until the situation took a turn for the worse in 1943, in Russia it was the complete opposite. When German forces invaded the Soviet Union, Russia was ill prepared and its forces were initially easily overrun. The Luftwaffe pilots were able to accumulate individual scores of hundreds of Soviet planes shot down. Russia needed a ‘wonder weapon’ quickly. The development of an operational rocket propelled interceptor became an instant priority. As the situation improved, military support for rocket plane development actually lessened in Russia, while at the same time in Germany the need for such aircraft rapidly increased.

The history of rocket planes in Russia starts with a club of rocket enthusiasts led by rocket engine developer Fridrikh Tsander called GIRD, the Russian abbreviation for ‘Group for Research of Reactive Propulsion’. Working without financial support, the members jokingly explain that the name of their organization actually means ‘Group of Engineers Working without Money’. Inspired by the Russian theoretician Konstantin Tsiolkovsky and the German rocket pioneer Hermann Oberth, Tsander is a prominent advocate of spaceflight but he doesn’t agree that other planets should be reached by expendable rockets. In an article ‘Flights to Other Planets’ published in the journal Tekhnika I zhizn in 1924 he explains that parachutes are less than optimal for landing a rocket on a planet which has an atmosphere, or indeed for returning a spacecraft to Earth, as parachutes do not offer the possibility of landing in a precise location. He continues: “For descending to a planet having sufficient atmosphere, using a rocket, as proposed by K. E. Tsiolkovsky, will also be less advantageous than using a glider or an aeroplane with an engine, because a rocket consumes much fuel during the descent, and its descent will cost, even if there is only one person in the rocket, tens of thousands of rubles, whereas descending with an aeroplane costs only several tens of rubles, and with a glider, nothing at all.” In 1933 GIRD manages to launch the first Soviet liquid propellant rocket (called the GIRD 10) to an altitude of 400 meters (1,300 feet). The rocket and its engine were mostly designed by Tsander, but sadly he died shortly before the test.

Another prominent GIRD member was Sergei Korolev, who would go on to become famous as the chief designer of the early Soviet space rockets. During the Cold War Korolev was the Russian equivalent of Wernher von Braun, and was responsible for the launches of Sputnik in 1957 and cosmonaut Gagarin in 1961 that began the space race and caused President Kennedy to direct NASA to beat the Russians to a manned landing on the Moon.

Shortly after the creation of GIRD in 1931, Korolev approaches Tsander with an idea for a rocket plane. He proposes to base this RP-1 on the existing BICh 8 tailless flying-wing glider and to power it using Tsander’s OR-2 rocket motor, which used gasoline and liquid oxygen and delivered a thrust of about 500 Newton. Later it is decided that the BICh 11 glider is better suited. In May 1932, Korolev becomes the director of GIRD and continues the design of the RP-1 and a successor called the RP-2. His work on the RP-1 and RP-2 designs evolves into the more ambitious RP – 218, a two-seat rocket plane for high-altitude research equipped with a pressurized cabin. A fixed main landing gear is planned initially, but this is soon changed to a retractable undercarriage. The RP-218 was to be taken to altitude by a carrier

Concept for Korolev’s RP-218.

airplane and released to boost itself to an altitude of 80 km (50 miles) or higher; to the edge of space, in fact. The plane is never built, partly because no rocket motor with the requisite thrust and weight is available. Interestingly, the envisaged mission scenario was very similar to that of the US X-15 rocket plane of the 1960s.

In 1935 Korolev designs the much simpler wooden SK-9 two-seater glider, which he intends to serve as a test bed for rocket motors. The RNII Rocket Scientific Research Institution, a new professional organization created by the merging of GIRD and the Gas Dynamics Laboratory in Leningrad in September 1933, sets out to transform the glider into a rocket plane by equipping it with an OR-2 engine. Being optimized to fly at relatively low speeds, modified gliders are useful for testing low-thrust engines such as this but they are not normally aerodynamically and structurally designed for the loads that come with powerful engines and high velocities. Otherwise gliders are well suited for conversion to rocket planes because they are fairly cheap and easy to modify, they do not have engines in the nose that need to be removed, they can be towed into the air by another airplane prior to starting the rocket engine, and they are able to glide to a safe landing once the propellants have been consumed. But great care has to be taken to ensure that the balance of the original design is not disturbed too much; hence most of the added weight has to be located near the original center of gravity. An added complication is that the weight of the propellant will decrease as the engine runs, whilst the fixed weight of the added structures, tanks and engine will remain unchanged. If this is not properly taken into account, an aircraft that is nicely balanced at take-off can become unstable as the flight progresses.

The rear seat of the SK-9 is replaced by a tank for 10 kg (22 pounds) of kerosene and two tanks for 20 kg (44 pounds) of fuming nitric acid. The rocket motor and its nitrogen pressurization system are installed in the aft fuselage, with the nozzle exit under the slightly modified rudder. The resulting RP-318 rocket plane has a take-off weight of 660 kg (1,460 pounds), a wingspan of 17 meters (56 feet) and a length of 7.4 meters (24 feet). The test phase begins with ground firing tests, initially with the OR – 2 controlled from the cockpit but separated from the plane by an armor plate (in case it blows up) and later with the engine installed in the aircraft. More than 30 test firings are performed. In April 1938 the plane is deemed ready for flight but then the notorious ‘purges’ of the paranoid Soviet leader, Josef Stalin, take their toll. Already, in 1937, the director and the chief engineer of the RNII institute (now called

NII-3) were executed. But a war with Germany is looming and Stalin finally realizes that shooting Russian aircraft engineers is detrimental to the quality of his air defenses. By locking them up in development center prisons instead, it will be possible to keep an eye on them while they help to design planes for the war effort. Valentin Glushko, the chief engine designer, is sent to the Butyrka prison and works in a special design bureau for “subversive elements”. When in June 1938 Korolev himself is declared an “enemy of the people” and sentenced to ten years’ hard labor in the horrible Kolyma gold mines in which temperatures regularly drop to minus 38 degrees Celsius (minus 36 degrees Fahrenheit), the development of the RP-318 halts. Although Korolev is soon transferred to a prison design bureau on the request of famous aircraft pioneer Sergei Tupolev (himself in prison) he is not permitted to work on rocketry except at night in his own time.

Only near the end of 1938 is the RP-318 project resumed at NII-3, now under the leadership of Arvid V. Pallo and without the involvement of Korolev. It is decided to repair and modify the existing prototype into the RP-318-1, involving the rebuilding of the tail section that was damaged during the ground tests, the installation of a new landing ski, and the fitting of a shock absorber to the tail skid. The most important change is the replacement of the OR-2 engine by the more powerful ORM-65 rocket engine designed by Glushko and Leonid Pushkin. But concern soon arises about this choice. The ORM-65 has been ground tested many times and flown nine times on the winged experimental RP-212 cruise missile designed by Korolev, but it is not really suitable for a reusable piloted aircraft. The engine can only be ignited on the ground and, once operating, cannot be turned off. Also the heat of the combustion degrades the engine to such an extent that it will be dangerous to attempt to use it for several flights. The design is therefore modified to make it compatible with a manned rocket plane. The resulting RDA-1-150 engine is 2 kg (4 pounds) lighter, has an improved cooling system and redesigned propellant injectors, can be ignited in flight in a low thrust regime in which only about 10% of the normal amount of propellant flows into the combustion chamber, and its operating thrust can be regulated between 700 and 1,400 Newton. The new engine is ground tested over 100 times, including 16 times while installed in the actual aircraft.

The flight test program with Vladimir Pavlovich Fedorov as pilot begins in early February 1940 at a grassy airfield in Podlipki near Moscow. It starts with unpowered glide flights using a dummy engine for mass balance and with the plane being towed to altitude by an R-5 biplane. On the first flight the propellant tanks are empty, on the second they are half filled, and on the third the full tanks are slowly drained to mimic propellant consumption by the engine. In February 1940 three low-power flights are conducted during which the engine is run in the low-thrust ignition regime. Fedorov reports that the engine can be clearly heard in his open cockpit, which means that its proper functioning can be monitored by its sound as well as by the instrumentation in the cockpit.

Finally the aircraft is ready for a full powered flight. On 28 February 1940 the RP – 318-1 is towed to an altitude of 2.8 km (1.7 miles) and released. It glides 200 meters (660 feet) down before Fedorov ignites the rocket engine. Gray smoke shows that the powder charge has fired to ignite the engine’s liquid propellant, then brown smoke shows that the engine is operating in its starting regime. As Fedorov further opens the propellant flow, a bright, almost smokeless flame nearly 1.5 meters (5 feet) long streaks from the nozzle. It pushes the RP-318-1 from 80 to 140 km per hour (50 to 90 miles per hour) in 5 seconds. Next, the plane climbs up to an altitude of 2.9 km (1.8 miles). When the engine stops after 110 seconds of continuous operation, Fedorov glides the aircraft to a safe landing. The plane makes another two successful powered flights on 10 and 19 March, after which the arriving spring melts the snow and turns the airfield into an unusable mire. In the autumn the plane is returned to NII-3 and dissembled. There are plans for further tests using another engine and a jettisonable wheeled dolly but these never take place due to other priorities within the institute. In August 1941 the German Army approaches Moscow, and as the institute prepares for evacuation to the safety of the Ural mountains the RP-318-1 is burned to prevent the Germans from finding it.

Meanwhile, in 1940 NII-3 makes a study of a plane primarily powered by wing – mounted ramjet engines and an RD-1400 rocket engine in the tail to propel the plane up to the speed required for the ramjets to start working. In the spring of 1941 a draft concept called the Tikhonravov 302 is put together under the leadership of Mikhail Tikhonravov. This is later approved by the director of the institute, Andrei Kostikov, and Tikhonravov leads a team of engineers in developing the detailed design. He also assumes responsibility for the necessary aerodynamic calculations.

Kostikov had become director of the institute after the execution of his predecessor and the imprisonment of Korolev and Glushko; all of which he seemingly engineered to advance his own career. When war breaks out with Germany and Stalin takes an interest in the new project, Kostikov decides to claim the design in the expectation that its success will raise his standing with Stalin. In November 1942 Stalin names him chief designer, and Tikhonravov’s revolutionary airplane concept becomes the Kostikov 302.

The Soviet State Defense Committee gives Kostikov only one year to get the 302 into the air. He duly promises that it will be able to fly for 20 minutes at an altitude of 8 km (26,000 feet) at a speed of 800 km per hour (500 miles per hour). A budget is allocated for the construction of two prototypes by NII-3’s OKB-55 experimental production facility headed by Matus R. Bisnovat, and a series of flight tests. The 302 is mainly wood but the elevators are made of aluminum alloy. The straight, tapered, low-slung wings have a shght dihedral for enhanced roll stability. It has a pressurized cockpit, an undercarriage with retractable main wheels and a retractable tail wheel, and hydraulic actuators to help the pilot to handle the expected large forces on the control surfaces (normally fighter planes of that time were operated by muscle power alone, which could make steering a plane at high speed with high aerodynamic forces very ‘heavy’). Because of the war, the test phase is to be minimized and the airplane turned into an operational fighter as soon as possible. Hence the 302 prototypes are equipped with armored glass in the canopy, an armor plate under the instrument panel and four ShVAK 20-mm cannon; two in the nose and two in the forward belly of the plane. In addition, the aircraft would be able to carry rockets on rails under its wings, or two 125 kg (276 pound) bombs for ground targets. The ramjets were to have been installed under the wings but development difficulties lead to the decision

to cancel them altogether and operate the plane by rocket power alone, as the 302P (‘Perekhvatchik’, Russian for Interceptor).

The lack of ramjets greatly reduces the potential range of the aircraft but it is still valuable as a short-range interceptor for fast, short-duration attacks on enemy aircraft (essentially giving it the same role as the German Me 163). The new engine chosen for the plane is the RD-2M-3V developed by Dushkin and Glushko. Like the engine developed for the German Me 263 (and for the same reason) it has two combustion chambers: a large one with 11,000 Newton of thrust for take-off and ascent, and a smaller 4,000 Newton chamber for more economical cruise flight (these numbers are the thrust given at sea level; at higher altitudes the thrust is slightly greater because atmospheric pressure hinders the outflow of the exhaust). The propellant load for this engine consists of 505 kg (1,110 pounds) of kerosene and 1,230 kg (2,710 pounds) of 96% concentrated nitric acid; a nasty and dangerous substance. These propellants are not hypergolic, so an igniter is required to initiate combustion. The pumps providing the propellant to the combustion chambers operate using 80% concentrated hydrogen peroxide and provide gas at high pressure by the process of decomposition.

In the spring of 1943 the first of the two prototypes built by OKB-55 is sent for testing in the large T-104 wind tunnel of the TsAGI institute. Glide test flights begin in August 1943 with the 302P prototype being towed by a Tupolev SB bomber to its release altitude. V. N. Yelagin is the engineer responsible for these test flights, with pilots Sergey N. Anokhin, Mark L. Gallai and Boris N. Kudrin (the nation’s oldest

One of the Kostikov 302 prototypes.

active test pilot). The twelve tow flights reveal serious stability problems at speeds over 200 km per hour (120 miles per hour), which is of course a major hurdle for a plane that is one day expected to fly at transonic speeds. Modifications are made and another series of glide flights are conducted in which the 302P is towed by Tupolev SB, Tupolev Tu-2, and North American B-25 bombers (the latter one of the planes donated by the US to help the Russians in their fight against Germany). This time the 302P proves to be exceptionally stable and rather easy to fly and land while gliding, so expectations for its handling under rocket power are high. However, development of the rocket engine is lagging far behind. In early 1944 it is still not able to reach the performance required to make the 302P the effective interceptor that Kostikov promised.

With the ram engines deleted, the plane is now expected to be able to fly at 8 km (26,000 feet) altitude for only 5.3 minutes instead of the specified 20 minutes, and at 725 km per hour (450 miles per hour); some 75 km per hour (50 miles per hour) less than originally specified. If the 302P were to fly at the original altitude and speed, it would run out of propellant in just 2.5 minutes. The absolute top speed also dropped from 900 km per hour (560 miles per hour) to 800 km per hour (500 miles per hour). The machine is still expected to be able to shoot up to 9 km (30,000 feet) in only 2.8 minutes but the military value of the plane as an interceptor is now highly doubtful. Moreover, whereas at the start of the project the Soviet Union was being overrun and overflown by the Germans, by early 1944 the need for an advanced interceptor plane is less urgent. In March 1944 it is decided to cancel the entire program, although one powered flight of the first 302P prototype remains in the planning. In the winter of 1944 it is fitted with skids and makes its first and final flight powered by an RD-2M3 rocket engine. Details are scarce but eyewitnesses say the flight was a success, even though one of the undercarriage ski legs failed at touchdown and caused the plane to slew into a snowdrift. However, the failure of the 302P as an effective weapon has dire consequences for Kostikov. Despite having the ‘Hero of Socialist Labor’ medal and the Stalin Prize, he is sent to prison on 15 March 1944 for obstructing the war effort and is not released until 1945.

During 1942 Aleksandr Yakovlev, the famous designer of Russian fighter aircraft, worked on the design of a concept very similar to the 302. It was to be based on his Yak-7, but instead of a piston engine and propeller it was to be driven by two ramjets mounted under the wings and a Dushkin D-l-A liquid propellant rocket motor in the rear fuselage. The result would be the Yak-7R (the ‘R’ standing for ‘Reaktivnyy’, meaning ‘Reaction-propelled’). But the project never left the drawing board owing to the lack of reliable ramjets.

The most famous Soviet rocket plane of the Second World War is the Bereznyak – Isaev BI. This story begins in the spring of 1940, when the Zhukovsky Institute in Moscow (TsAGI) hosted a conference on ramjet and rocket propulsion. It is probable that the meeting was inspired, at least partially, by the advanced rocket and aircraft developments in Germany, which they knew of because Soviet intelligence had been able to recruit Willy Lehmann, a German Gestapo officer who kept them informed. Among the attendees were Viktor Fedorovich Bolkhovitinov, head of design bureau OKB-293, and Aleksandr Ya. Bereznyak and A. M. Isaev, two of his top engineers. Both Bereznyak and Isaev were very excited by the idea of rocket propelled aircraft and convinced Bolkhovitinov to let them start to design one. In the autumn of 1940 they showed fellow engineer Boris Chertok a preliminary design of what they called ‘Project G’. It was a concept for a compact plane built of wood and duralumin (an aluminum alloy) with a take-off weight of 1,500 kg (3,300 pounds). To serve as an operational interceptor it was to have four machine guns; two 12.7­mm and two 7.6-mm caliber. The engineers intended to power their plane with a new

14,0 Newton rocket engine burning low-grade kerosene and red fuming nitric acid that was under development at NII-3 by a team led by Leonid Dushkin (who had also designed the RDA-1-150 that powered the RP-318-1). Since the thrust of the engine was close to the maximum weight of the plane, which would decrease as it burned propellant, it would be able to climb almost vertically once airborne. The top speed was estimated at 850 to 900 km per hour (530 to 560 miles per hour). According to the designers the most important selling points were the incredibly fast climbing rate (enabling it to reach enemy planes quickly and intercept them by surprise), its high speed (making it virtually invulnerable to enemy fighters), and the inherent simplicity of the rocket in terms of manufacturing and maintenance when compared to high-performance piston engines.

Bereznyak, Isaev and Chertok visited NII-3 in March 1941 to check on the status of the rocket engine, called the D-l-A-1100. This was state-of-the-art at the time. It weighed 48 kg (106 pounds) and consisted of several large forged-steel sections (the conical head with 60 propellant injectors, the cylindrical combustion chamber and the nozzle) joined using bolts and copper gaskets. Cooling was provided by pumping the kerosene fuel through the double-walled combustion chamber and the nitric acid oxidizer through the double-walled nozzle. The engine was ignited using a glowing plug of nichrome; later replaced by silicon carbide. But this marvelous technology was not working yet, mostly due to problems with Dushkin’s innovative turbopump driven by hot gas and steam produced by a small combustion chamber that was fed a mixture of water and the same propellant as the main rocket combustion chamber; an efficient but complicated affair. Furthermore, the engine was not going to deliver the specified 14,000 Newton of thrust: the prediction was that it would deliver no more than 11,000 Newton. On 21 June Isaev proposed running the engine’s turbopump on compressed air instead; it would result in a heavier rocket engine but would be much simpler.

The very next day Germany invaded the Soviet Union and the need for the rocket propelled interceptor instantly became very urgent. Bereznyak and Isaev set to work on a new, more detailed design. They finished it in only three weeks and on 9 July, together with Bolkhovitinov, met with the head of NII-3, the earlier mentioned Andrei Kostikov. Although Dushkin was not happy with the idea of altering his fuel pump, Kostikov agreed that the urgency of the situation meant that using compressed air made sense. A letter was sent to the Kremlin, and it was even shown to Stalin personally. The Project G team went to Moscow to report on their design and were ordered to build the plane in only 35 days. The engineers were given leave to visit their families, then literally lived in the factory to meet the extremely demanding deadline. The same order tasked NII-3 to finish the development of the D-1-А-1100 engine as soon as possible, and make it capable of multiple restarts in flight as well as thrust variation in the range 4,000 to 11,000 Newton.

The new design was called the BI for ‘Blizhnii Istrebitel’ (Close-range Fighter) and also the first letters of the two inventors Bereznyak and Isaev, although whether this was deliberate is unclear. It was now a sleek low-wing machine with a length of only 6.4 meters (21 ft) and a wingspan of 6.5 meters (21 ft). The take-off weight was 1,650 kg (3,640 pounds), of which 710 kg (1,570 pounds) was propellant. This made the BI a very diminutive fighter aircraft; smaller and lighter than the Me 163B. For comparison the conventional German Messerschmitt Bf 109G fighter had a length of

9.0 meters (29 ft), a wingspan of 9.9 meters (33 ft) and a maximum take-off weight of 3,400 kg (7,500 pounds), and was still regarded as being a relatively small fighter. The BI was kept as simple as possible so that it could be produced in short order and in sufficient numbers to overcome the German invasion of Soviet airspace. The four machine guns planned earher were replaced by two more powerful 20-mm ShVAK cannon. The BI would have a wooden frame and a 2 mm (0.08 inch) plywood skin covered by a bonded fabric, and it would be easy to mass produce. The wings were to be relatively short to limit drag whilst still providing sufficient lift at the planned high flight speeds (lift is a function of both the area of the wings and the flight speed, so at higher speeds smaller wings are sufficient). However they were not particularly well designed for transonic flight phenomena (something that will result in a serious accident, as we shall see). The ailerons, elevators and rudder were covered by fabric but the flaps were duralumin. Ten tanks with compressed air (held at a pressure of 60 atmospheres), five in the forward section and five in the aft section, were required for the rocket engine’s turbopump, to retract and deploy the landing gear, and to power the cannon. The forward section also housed two kerosene fuel tanks while the aft section had three tanks of nitric acid oxidizer. The air tanks were made from a high – strength steel (Chromansil) that was great for making light pressure vessels but not very resistant to corrosion. Their proximity to the extremely corrosive nitric acid was thus rather hazardous and required the acid tanks to be replaced periodically in order to ensure that no leaks could develop.

Working around the clock (with local furniture makers supplying the wood-and – fabric airframe) the team delivers the first prototype on 1 September 1941. A second prototype is also being assembled. However, Dushkin’s engine is still not available. The first prototype, BI-1, is towed into the air by a Pe-2 bomber on 10 September with test pilot Boris Kudrin at the controls. Following release, he glides back to the airfield and makes a successful landing. Another fourteen unpowered flights follow and establish that the plane behaves well at low speeds. Interestingly, rival aircraft designer A. S. Yakovlev had the prototype towed to the TsAGI T-104 wind tunnel for testing. This alarmed the BI team because Bolkhovitinov had a rather rocky history with Yakovlev, but Yakovlev himself, and his aircraft designer, Ilya Florov, studied the results and gave the team good suggestions for improvements. In this way a yaw instability was corrected by enlarging of the rudder and adding two circular vertical plates at the tips of the horizontal stabilizer.

In addition to the problems with the engine, the BI project was delayed by the evacuation of both OKB-293 and NII-3 in October 1941 (in preparing for which, as mentioned above, the RP-318-1 rocket plane was burned). Most of Moscow’s vital war industries were moved deep into the Ural mountains to ensure they would not be overrun by the rapidly advancing Germans. The BI team was stationed in Bilimbay but Dushkin’s team ended up 60 km (37 miles) away in Sverdlovsk. Near their new accommodation the team built a test stand for their aircraft on the shore of the frozen Lake Bilimbay. It comprised a cradle that could hold the plane during engine tests, as well as measure the thrust. But there was still no engine to install. Dushkin was now increasingly absorbed by other projects (including NII-3’s own Kostikov 302 rocket plane) but he assigned his engineer Arvid Pallo to oversee the installation and testing of the rocket engine. When the static test campaign finally begins in early 1942 it becomes immediately clear that the nasty nitric acid is trouble: it corrodes parts of the airplane as well team members, causing skin burns and respiratory irritation. Tanks of sodium carbonate solution have to be kept handy to neutralize the all-too-frequent acid spills. For these tests the new test pilot Grigory Yakovlevich Bakhchivandzhi (Kudrin was ill) operates the engine from the cockpit: a method fraught with risk, as we have seen with the He 112 which almost killed Erich Warsitz during a ground test. Indeed, on 20 February the BI-l’s engine explodes, blasting the nozzle section into the lake. The forward assembly of the engine smashes through the airframe and strikes the rear of Bakhchivandzhi’s seat, knocking him against the instrument panel. Fortunately his injuries are minor. Nitric acid spraying from a broken propellant line drenches Pallo. Luckily his eyes are saved by his protective glasses and the rest of his face is partly spared by alert mechanics who dunk him head-first into a tank of soda solution. His scars serve as a grim reminder of the dangers of rocket testing. A study of the engine debris reveals that the combustion chamber had succumbed to corrosion fatigue. In pushing on with the testing phase despite a shortage of engines, the one D-l-A-1100 available had been operated too many times and for too long. After the test bench is rebuilt and the engine’s propellant supply system improved, static firing tests resume, with the undeterred Bakhchivandzhi performing three of them. A 5.5 mm (0.22 inch) steel plate is added to the rear of the pilot’s seat as protection.

By April 1942 the BI-1 is deemed ready for flight testing at the nearby Koltsovo airfield (now the main airport of the city of Sverdlovsk). Interestingly, in spite of the obvious danger, the nitric acid for the engine is actually transported to the airfield in glass bottles and there poured into the containers from which it will subsequently be fed into the plane’s tanks. All of this is done without any protective clothing.

On 2 May Bakhchivandzhi takes the controls of the BI for a short, low-thrust hop one meter above the ground. Then on 15 May he prepares for the first real test flight. Like the Me 163, the BI could explode if it were to make a hard landing before all the propellant was consumed, so it is loaded with only 240 kg (530 pounds) of nitric acid and 60 kg (130 pounds) of kerosene. He arrives wearing a new leather coat and shiny boots but when he climbs into the cockpit he is wearing his old flying gear: he explains that if he doesn’t survive the flight the new coat and boots will be useful for his wife… she could make some money by selhng them! Taking off under rocket power, he leaps into the air and quickly reaches an altitude of 840 meters (2,800 feet) and a speed of 360 km per hour (220 miles per hour) with the thrust limited to 8,000 Newton. Another safety measure, a fairly standard one for the early flights of new airplane prototypes, is that the undercarriage is kept down. After one minute, an indicator lamp reports the rocket is overheating, so he shuts the engine off. Gliding in to land, the plane descends too rapidly owing to insufficient forward speed and the resulting lack of lift. The touchdown is hard and breaks the undercarriage but the plane is not significantly damaged. Bakhchivandzhi lives to wear his new coat and boots another day.

Compared to today’s test flying, taking up a plane for the first time was in those days a truly scary business, especially if it was powered by a rocket engine. Modern airplane prototypes have already been exhaustively tested by sophisticated computer simulations long before metal is cut. The designs are based on an enormous database on subsonic, transonic and supersonic aerodynamics, control theory, materials and so on. Engines are tested hundreds of times before being deemed sufficiently reliable to be fitted into an airplane. All of this ensures that the question to be answered by the first flight is not whether it will fly, but rather how well it matches the performance predictions.

In contrast, pilots taking up revolutionary barely tested machines like the Me 163, Syusui and BI were really pushing very experimental technology into the unknown. Rather than wondering how the planes would fly, their minds were probably more occupied with the question how these new beasts were going to try to kill them. And unlike modem pilots, these pioneers did not have teams of engineers monitoring the flight on the ground, ready to help in case of trouble; once in the air, they were truly on their own. Last but not least, they had no ejection seat to instantly boost them out of a hairy situation.

The BI’s first flight lasted only 3 minutes and 9 seconds but was judged a success, with Bakhchivandzhi reporting, “the aircraft performed stable decelerations, gliding and handhng like any ordinary aircraft”. The State commission in charge of assessing the BI project delightedly noted: “The take-off and flight of the BI-1 aircraft with a rocket motor used for the first time as the aircraft’s main engine has proved the practical feasibility of flight based on a new principle; this opens up a new direction in the development of aviation.” This flight is sometimes hailed as that of the world’s first operational rocket fighter plane, but this is somewhat of a stretch. Even though the BI was designed as an operational weapon, it was still very much experimental. It is certainly true that it flew before the first Me 163B in Germany, but the DFS 194 and Me 163A predecessors of that machine had taken to the air long before the first BI flight.

Encouraged, Stalin authorizes the production of a batch of 30 BI interceptors for operational military service. These are enhanced with racks for small bombs that can be dropped onto enemy bomber formations by flying above them, causing damage by a combination of shock waves and shrapnel. The decision to declare the BI ready for military service after only a single low-speed flight soon proves to be premature.

In July, Pallo is recalled to NII-3 by Dushkin to help him with the institute’s own Kostikov 302 rocket plane. Isaev takes over management of the BI’s rocket engine in OKB-293. To help get started, he goes to learn the tricks of the trade from Valentin Glushko in the Butyrka prison design bureau. Glushko shows Isaev how to improve the engine, and Isaev starts to work on what will become the RD-1 rocket engine that will later be incorporated in the BI aircraft.

After its first and only flight, the BI-1 is deemed too damaged by nitric acid spills for safe flight and it is retired. Testing continues using the second prototype. On 10 January 1943 Bakhchivandzhi flies the BI-2 for the first time. He still limits the D-l – A-l 100’s thrust to 8,000 Newton and the speed to 400 km per hour (250 miles per hour) and hence achieves an altitude of only 1,100 meters (3,600 feet). This time the undercarriage, with skids instead of wheels for taking off from the snowy airfield, is retracted, resulting in a smoother flight with aerodynamics more representative of a BI interceptor in operational use.

The second BI prototype.

A third flight (the second with the BI-2) is made two days later by another pilot, Konstantin Gruzdev. (Bakhchivandzhi was at that time visiting NII-3 to check upon progress with the Kostikov 302; there was clearly a great deal of interaction between the rocket plane teams of NII-3 and OKB-293.) This time the propellant flow is fully opened, allowing the engine to provide its full thrust of 11,000 Newton. At the peak altitude of 2,190 meters (7,190 feet) Gruzdev achieves a speed of 675 km per hour (420 miles per hour). When he extends the undercarriage one of the skids breaks off but he manages to land safely. Asked about the flight, Gruzdev comments, “It’s fast, it’s scary, and it really pushes you in the back. You feel like a devil riding a broom.” Henceforth the ‘devil’s broomstick’ nickname is used by everyone who works on the BI project. Film of this flight has survived, showing the BI accelerating quickly over the frozen lake, taking off and shooting up. The less than perfect landing was also captured, with the plane toppling forward because of the broken skid and performing a ground loop as it skidded across the ice. If such a thing had happened on a grassy airfield the plane would probably have burrowed its nose in the ground, with a much more severe outcome.

On his return Bakhchivandzhi takes over from Gruzdev and flies the refurbished ski-equipped BI-1 prototype on 11 and 14 March 1943, and on the 21st he takes off in the third BI prototype (also on skids) and climbs at a maximum rate of 83 meters (272 feet) per second; this is about half the rate of an Me 163B but still at least five times better than contemporary Russian piston-engine propeller fighters. On the 27th Bakhchivandzhi makes another low-altitude test flight. After 78 seconds, as he opens the throttle to push the plane beyond its speed record, the BI-3 suddenly goes into a 50 degree dive and smashes into a frozen lake instantly killing the pilot. The BI team has discovered the infamous transonic ‘Mach tuck’ phenomenon the hard way. This is confirmed by tests in TsAGI’s new T-106 high-subsonic wind tunnel. The BI – 3’s onboard recording instruments were too badly mangled by the crash to give accurate data on the final speed, but estimates range from 800 up to an astonishing 990 km per hour (500 to 620 miles per hour). This disaster prompts the Air Force to cancel the pre-production order. For his achievements and ultimate sacrifice Bakhchivandzhi is posthumously awarded the ‘Order of Lenin’. (In 1973 he gained the more prestigious ‘Hero of the Soviet Union’.)

In May 1943, with OKB-293 relocated to Moscow following the German retreat, Bolkhovitinov writes a detailed report on the experiences with the BI prototypes. He emphasizes the need to study the shock effects that had caused the BI-3 to crash, and in order to investigate transonic and supersonic flight dynamics he recommends the development of a rocket plane capable of 2,000 km per hour (1,200 miles per hour).

No BI flights are made for over a year, probably in part because the urgent need for an operational rocket interceptor has lessened considerably, but also because time is required to build new airplanes. In 1944 five more prototypes are readied. BI – 6 is given a pair of ramjet engines instead of a rocket engine. It is towed into the air on three occasions but the test pilot never manages to get the engines to work properly. The other prototypes are to be fitted with the new RD-1 rocket engine designed by Isaev, the first example of which is completed and tested in October 1944 (it should not be confused with the smaller RD-1 engine of Dushkin and Glushko).

BI prototype number 5.

The general layout of the engine is similar to that of the D – 1-А-1100 and also has a maximum thrust of 11,000 Newton, but Isaev has introduced numerous improve­ments, including a more reliable electric-arc igniter instead of the glow plug, and new injectors designed to improve the fuel and oxidizer mix in the combustion chamber. The BI-7 is fitted with the Isaev RD-1 and flown by test pilot Boris Kudrin on 24 January and 9 March 1945. The maximum speed attained is 587 km per hour (365 miles per hour), which is well short of the dangerous transonic flight regime. These flights reveal a problem of excessive vibration in the tail. Gliding tests using the BI-5 and BI-6 are made to investigate the problem but the pilots are unable to recreate the vibration. However, by this point the war is nearly over and there is no operational requirement for the BI interceptor, so work is halted. The ‘devil’s broomstick’ is never flown in combat but its developers have gained valuable experience that will be put to good use after the war in developing new rocket propelled aircraft.

No examples of the BI survive but a reasonably accurate replica can be seen in the Russian Federation Air Force Museum in Monino. At the airport of Sverdlovsk, a monument of a BI replica shooting into the sky commemorates the rocket aircraft’s first flight from that location.

Another rocket plane that however never progressed as far as the BI was the ‘Malyutka’ (Little One), the final aircraft designed by famous aviation pioneer N. N. Polikarpov. The construction of a prototype started in early 1944 and was nearly complete by the middle of the year. However, all work was suddenly stopped when Polikarpov died of a heart attack on 30 July of that year, and his facilities were absorbed into the rival Lavochkin design bureau. Work on the project was never resumed.

The Malyutka was initially planned to use a D-l-A-1100, the same rocket engine as powered the BI, but it was later decided to use a dual-chamber RD-2M-3V, which was the same engine as intended for the Kostikov 302P. The combined thrust of the two chambers was expected to be sufficient for a top speed of 845 km per hour (525 miles per hour). As the propellants were consumed, the changing center of gravity of the aircraft would cause stability problems. The solution was to equip the plane with a tank from which water could be discharged in order to compensate for the changing weight balance; simple but not very elegant. The fuselage would be made of wood, but unlike any previous Polikarpov fighter the wings and tail section would be lightweight aluminum alloy. The aerodynamic control surfaces would be operated by a pneumatic system, the cockpit would be pressurized for high-altitude flight, and the armament would comprise two powerful VYa-23 23-mm cannon. In contrast to the conventional undercarriage of the 302P and the BI, the Malyutka would employ an undercarriage and a retractable nose-wheel instead of a tail wheel. A plane with a tail wheel (a so-called ‘tail-dragger’) has its nose angled up whilst taxing, which makes such an undercarriage a good choice for a machine driven by a propeller that has to clear the ground. However, a jet or rocket aircraft has no need for a large ground clearance of its nose and so can be equipped with a nose wheel in order to align the plane more horizontally on the runway. This gives the pilot a much better view of where he is going during taxiing and the take-off run. Furthermore, with the plane in a horizontal position there is less danger of jet or rocket engine exhaust melting the runway. While most propeller fighter aircraft of the two world wars had tail wheels, nearly all jet fighters employ so-called tricycle undercarriages, with wheels under the wings and the nose.

The idea of adding rocket engines onto existing piston-engined airplanes (like the Germans with the Heinkel 112) was also explored as a stop-gap while waiting for the rocket and jet airplanes. The 3,000 Newton thrust Dushkin and Glushko RD-1 engine was fitted to various types of aircraft to boost their performance. On 1 October 1943 tests began with the engine in the aft fuselage of a Petlyakov Pe-2 dive bomber, with the turbopump driven by one of the two standard piston propeller engines. The tests with the Pe-2RD prototype revealed problems with the RD-l’s electrical ignition, so it was replaced by a chemical ignition system. Both the RD-1 and the RD-lKhZ (the improved engine) were also flown on two Lavochkin La-7 fighters (the La-7R1 and La-7R2), an older La-7 prototype airframe (the La-120, converted into the La-120R), a Sukhoi Su-7 high-altitude interceptor (the original Su-7, not the jet fighter that was introduced in the 1950s with the same designation) and a Yakovlev Yak-3.

These tests proved the concept. For instance, on 11 May 1945 the rocket-equipped Yak-3, the Yak-3RD, reached a speed of 782 km per hour (486 miles per hour) at an altitude of 7.8 km (25,600 feet); some 130 km per hour (80 miles per hour) faster than the top speed of a conventional Yak-3. Indeed, it seems that a La-7R with an operating rocket motor successfully participated in the Moscow air displays of 1946 and 1947. However, on several flights the RD-1 exploded as kerosene fuel and nitric acid oxidizer came into contact outside the combustion chamber owing to leaks. The brave test pilots were usually able to land their damaged planes. One pilot of the sole rocket propelled Su-7 prototype was not so lucky, because while preparing the plane

The Sukhoi Su-7 with a tail-mounted rocket engine.

for the first post-war air display over Moscow in 1945 the rocket motor exploded, destroying the aircraft and kilhng him. On 16 August 1945 test pilot V. L. Rastorguev died when his experimental Yak-3RD crashed for unknown reasons.

It became clear that apart from reliable and safe rocket engines, Russia also lacked vital knowledge on the dynamics of transonic flight. By 1935 Aleksandr Moskalyov had already drafted a concept for a rocket propelled aircraft that he thought should be able to exceed the speed of sound. The planform for this plane was based on that of his SAM-9 ‘Strela’ (Arrow), a propeller aircraft that had a revolutionary ogival delta wing (also known as a Gothic delta because its shape resembles the arches in Gothic cathedrals). Moskalyov expected this type of wing to be well suited for transonic and supersonic flight, and this was later confirmed by the supersonic, ogival delta-winged British-French Concorde and Russian Tupolev 144 (‘Konkordski’) airliners. Piston engines and propellers were not going to show the full potential of his wing, but with a rocket engine he expected to be able to reach transonic speeds and beyond.

With Dushkin’s help for the propulsion part, in 1944 Moskalyov came up with the design for the SAM-29, also known as the RM-1 (for ‘Raketnyi Moskalyov’, Russian for Moskalyov Rocket). Like the Strela, it had an ogival delta wing and big vertical stabilizer, but no horizontal tail. The planned engine was Dushkin’s RD-2M – 3V. To comply with the military’s rather impractical demand that any new rocket plane must be armed to serve as an operational fighter, the experimental RM-1 would have two cannon in the nose. Unfortunately, at the end of the war the project

was deemed too futuristic and in January 1946 Moskalyov’s design bureau was closed. Had those in power understood the RM-l’s potential and continued their support, then either it or a close descendant might well have become the first aircraft ever to fly faster than the speed of sound.

More or less in parallel with the RM-1, the development of another dedicated research rocket aircraft concept started in 1943. Designer Ilya Florentyevich Florov led the project for the Russian Air Force, and his Florov 4302/4303 was a relatively small rocket plane made entirely of light alloy. Its exterior had a very smooth finish to minimize aerodynamic drag, but unlike the Me 163 and various other high-speed German designs it had straight wings rather than swept back wings or delta wings. It is unlikely that the benefits of swept wings for transonic flight were fully understood in Russia at that time. (Nor indeed, as we shall see in the description of the post-war X-l, was the concept of the swept wing understood in the United States). The fully horizontal wings of Florov’s design, which were set high on the fuselage, had down­ward angled ‘flippers’ at the tips. These effectively produced a negative dihedral in order to avoid ‘Dutch roll’, a stability problem common to high-winged aircraft that imparts an out-of-phase combination of ‘tail-wagging’ and rocking from side to side. (The German He 162 jet plane that flew near the end of the war also had a high wing and similar drooping wingtips, which the Germans called ‘Lippisch Ears’.) The pilot was housed in a small pressurized cockpit. Three aircraft were built. The first had a fixed undercarriage with a tail wheel (using parts from a Lavochkin La-5FN fighter plane) and was intended for low-speed gliding flights only. The other two aircraft were for powered flight and (like the German Me 163B) were to take off employing an ejectable tricycle dolly, then land on a skid and a tail wheel. Aircraft 2 would be a

Florov 4302 (top) and 4303 (bottom).

Florov 4302 with a nitric acid/kerosene RD-1 rocket engine designed by Isaev with a maximum thrust of 11,000 Newton at sea level. Plane number 3 would be finished as a Florov 4303 with the RD-2M-3V two-chamber engine (the same engine that was planned for the Kostikov 302P and the Polikarpov Malyutka) delivering a combined maximum thrust of 15,000 Newton at sea level.

However, the flight test phase only begins in 1946, well after the end of the war. Pilots A. F. Pakhomov and I. F. Yakubov are appointed and that year make 46 towed glide flights with aircraft number 1. After some taxi tests and a short hop, Number 2 is flown for the first time under rocket power in August 1947, with Pakhomov at the controls. He is towed to an altitude of 5.0 km (16,400 feet) by a Tu-2 bomber. After release, he ignites the rocket engine and quickly accelerates to a speed of 826 km per hour (513 miles per hour), which is rather daring in a new experimental airplane that has not previously been tested under power at lower thrusts and speeds. Afterwards Pakhomov reports that the plane behaved well in all respects, and there were no vibrations. Several more flights with the 4302 then follow, and on one occasion a propellant feed line ruptures and noxious acid vapors slightly intoxicate the pilot. But by late summer 1947 the concept of the Florov 4302/4303 was obsolete thanks to the new information on high-speed flight and aerodynamics obtained from the defeated Germans. Moreover, jet fighters were now reaching similar speeds as those for which the 4302/4303 were designed. It was therefore decided that sufficient data had been gathered from the testing and that it would be better to concentrate effort and funding on the more advanced MiG 1-270 rocket plane (described in the next chapter). When the 4302/4303 program was halted, aircraft number 3 was still awaiting its engine.


While on the subject of future spaceplanes let us have a look at how these were (and still are) depicted in fiction. The existence of spaceplanes as an efficient means of transportation between Earth and space is generally taken for granted in science fiction, much like airliners are used in the real world. Apart from ideal aircraft-like operations, one thing that most of the spaceplanes in science fiction books, movies and television series appear to have in common is an amazingly efficient propulsion system because most of the volume is available for passengers and cargo rather than being taken up by bulky propellant tanks.

Take for instance the Orion III ‘Pan Am Space Clipper’ of the famous movie 2001, A Space Odyssey of 1968. We are shown a spaceplane that operates much like an airliner, even with stewardesses. No take-off is shown in the movie but it appears that the Orion is a single stage vehicle with double-delta wings (similar to those of the Space Shuttle, which had not yet been designed when the movie was produced). Windows cover only a short stretch of the middle of the fuselage, which makes sense for a spaceplane requiring large volumes of propellant. Still, as a relatively small single-stage-to-orbit spaceplane (judging from the size of the windows) 200Гs Orion must have fantastically efficient engines; jet engines, if the series of holes that can be seen on the leading edges of the plane’s wings and in front of the engine module are air intakes. However, having intakes on the wing leading edges would be rather poor engineering: tapping off the air before it has a chance to flow over the wings would seriously reduce lift. Nevertheless, Orion is probably the most believable spaceplane ever depicted in any block-buster movie.

The most famous space fighters of the movies, the Star Wars’ X-wing and Battle Galactica’s Viper, have wings and very apparent air intakes but spend most of their time in the vacuum of space. Even so, they seem to require only very small amounts of propellant as there are no large tanks to be seen. This does not seem to make sense: if you can fly and maneuver in space for hours without any external airflow, why bother with air intakes and wings for the brief periods in an oxygen-rich atmosphere? Moreover, these vehicles seem to maneuver in space as if they were flying through air, banking into nice round turns, flying in loops etc. In space there are of course no aerodynamic lift and drag forces, making such maneuvers only possible with the help of a serious set of reaction control thrusters which aren’t apparent in these fictional designs. Moreover, such airplane-like actions are rather useless in space (but not in a movie, as it certainly looks more exciting than real orbital mechanics).

Fireball XL5, a spaceplane shown in the British 1962-1963 television series of the same name also lacks apparent propellant tank volume, but according to the series it is powered by a “nutomic reactor” that enables it to have a range of many lightyears. Although the Fireball is a fictional vehicle without any regard for the limitations of real spaceflight, it is interesting to note that it takes off using a rocket powered sled and a mile-long launch rail that culminates in a 40 degree “sky ramp”, whereupon the spacecraft uses its own propulsion to ascend into space; very similar to the take­off mode of Sanger’s Silverbird.

In reality, even although serious work on orbital spaceplanes has been underway since before the Second World War, they still only fly in the realms of fiction. Like most of the hardware depicted in 2001, A Space Odyssey, nothing like an Orion III was available in the real year 2001; nor in 2011 for that matter. Only the videophone used the movie is actually better in today’s reality. In general, science fiction often overlooks the complexity of spaceplane transportation, and especially the limitations of propulsion technology. Of course, with engines working on fantasy and built out of unobtainium, anything is possible. What the spaceplanes of the movies do show us, however, is the ultimate goal of the real-world launch business: airline-like flights into space.


While engineers in Germany, Russia and Japan were busy building experimental rocket propelled interceptors and designing concepts for even more advanced rocket aircraft, very little work was done on this subject in the United States and the United Kingdom. The success of Allied conventional fighter planes in gaining air superiority over European and Japanese territories gave no incentive to investigate the potential of rocket planes.

The first rocket research to receive financial assistance from the US government was the development of solid propellant RATO (Rocket Assisted Take-Off) units for aircraft by the Rocket Research Group of the Guggenheim Aeronautical Laboratory at the California Institute of Technology (GALCIT). On 12 August 1941 a tiny Ercoupe sports plane piloted by Captain Homer A. Boushey Jr., was fitted with such a solid propellant booster and launched from March Field in California. The booster burned for 12 seconds and gave a thrust of only about 130 Newton. In a later test the RATO unit exploded during level flight. The Ercoupe made a safe landing, but there was clearly something not entirely right with the propellant. On 16 August, Boushey nevertheless took off in the Ercoupe with six RATOs firing. The next step was to get the plane airborne by rocket thrust alone, and this was achieved on 23 August. The propeller was removed and 12 RATO units were fitted. The little plane was towed to a speed of about 40 km per hour (25 miles per hour) by a truck and then the rockets were ignited. Although only 11 units actually fired, the Ercoupe left the ground and reached an altitude of about 6 meters (20 feet).

After these flight tests the team discovered that RATO units tended to explode if they were stored for several days instead of being used immediately after production. The propellant grain shrank and caused cracks and openings to develop between the propellant and the casing, and these caused sudden destructive surges in the internal pressure. Once it became clear that this problem could only be solved by using a new type of propellant, the team developed an innovative solid propellant which was a paste created by mixing black gunpowder with common road asphalt. Its mechanical properties were much more stable under various storage conditions than those of the earher powder propellant. Another major advantage of this composite propellant was that it could be cast into a predetermined shape, allowing propellant grains with pre-programmed thrust-over-time profiles (more or less equivalent to throttling a liquid propellant engine). This provided the basis for the sohd propellants for all later large solid propellant rocket motors, including those on the Space Shuttle.

Based on the new propellant (called GALCIT 53) the team managed to deliver on a Navy contract for 100 RATO units capable of prolonged storage in hot deserts as well as arctic conditions and then delivering a thrust of 900 Newton for 8 seconds. Shortly thereafter, production of operational units for the Navy began at the Aerojet

Take-off of America’s first RATO-equipped airplane on August 12, 1941 [NASA-JPL].

Engineering Corporation, a company set up by the GALCIT rocket research team for the commercial exploitation of their work.

Surprisingly, the only real American rocket aircraft under development during the Second World War was not an interceptor such as the German Me 163 or Japanese Syusui, nor even an experimental research aircraft like the Russian Florov 4303, but more like the rather desperate German and Japanese back-against-the-wall, last- stand designs for ramming aircraft. In 1942 John K. Northrop, the famous aircraft designer with a fascination for ‘flying wings’, came up with the idea for a fighter which would be sufficiently sturdy that it could slice right through an enemy bomber. His ‘Flying Ram’ was a tailless flying wing built like a knife, with a reinforced leading edge over most of its span and no vulnerable air intakes. Northrop intended it to be powered by an Aerojet XCALR-2000A-1 ‘Rotojet’ liquid propellant rocket engine (which did not yet exist) delivering a thrust of 9,000 Newton. Its take-off was to be assisted by a pair of 5,000 Newton thrust solid rocket boosters which would be dropped once airborne. The engine would run on mono-ethylaniline fuel and red fuming nitric acid oxidizer; a combination that we would not nowadays consider suitable for a manned military aircraft because it is rather toxic. Moreover this fuel would be especially corrosive to the innovative magnesium alloy structure which was to make the aircraft especially sturdy and resistant to damage. Northrop wanted the pilot to lie prone in the cockpit, in the expectation that this would enable him to survive the violent collision with an enemy plane. Moreover, this would make the flying wing much flatter so that the 8.5 meter (28 feet) wingspan would, seen from the front, present a minimum silhouette that would be difficult to hit by gunners in the targeted bomber and also enable it to more easily slice through the bomber.

Northrop managed to interest the US Army Air Forces (USAAF, forerunner of the US Air Force) in the project although at that time there was no obvious requirement for such a radical defense against enemy bombers. Perhaps intelligence forewarning of the secret long-range bomber developments underway in Germany was behind the support for this wild idea. In January 1943 the USAAF issued the Northrop company a contract for three rocket propelled prototypes under the designation XP-79, but in March it was decided to equip the third prototype with a pair of Westinghouse 19-B turbojets as the XP-79B. Similar to the development logic implemented in Germany, Japan and Russia, ghder test vehicles were built to verify the aerodynamics of the aircraft design. Two prototypes (designated MX-334) were pure gliders but the third (designated MX-324) was fitted with an Aerojet XCAL-200 rocket engine that used mono-ethylaniline and red fuming nitric acid. Its thrust of 900 Newton was rather puny compared to the 17,000 Newton of the engine that was at that time propelling the Me 163B in Germany, and a clear indication of how far US rocketry was lagging behind the developments in Germany, Russia and Japan. The airframes of the gliders consisted of a center section of metal tubing covered with plywood, wooden wings, and a fixed tricycle undercarriage.

Flight tests of the MX-334 glider, towed into the air by a P-38 Lightning fighter, showed the flying wing to be rather dangerously unstable even after the addition of a simple vertical fin that was held in place using metal cables. On one flight, the glider got caught in the propeller wash of the tow plane just after release, which caused it to suddenly pitch up, stall and enter a spin. Test pilot Harry Crosby managed to recover from the spin but ended up flying inverted. Finding himself lying on the roof of the cockpit unable to reach the controls, he managed to open the hatch and parachute to safety. The MX-324 was tested at Harper Dry Lake in the Mojave Desert of southern CaUfornia, where secret airplanes by both the Hughes and the Northrop companies were tested. On 20 June 1944 ground tests of the rocket motor were started, followed by taxiing tests with Crosby at the controls. On 5 July the engine and the plane were deemed flight-ready. He eased himself into a prone position in the cramped cockpit with his head in a sling to enable him to look ahead through the large windshield. In front of him he saw the P-38 that was going to tow him to release altitude. The two planes took off and climbed up to 2.5 km (8,000 feet), where Crosby triggered the towline release, then braced himself as he pressed the engine ignition button on the control stick. The thrust was very modest but still resulted in an acceleration of 0.08 G due to the plane’s low total mass of about 1,130 kg (2,500 pounds). When the engine ran out of propellant after 4 minutes, Crosby glided the plane down to a gentle landing on the dry lake. Thus the MX-324 became the first true US rocket propelled aircraft to fly (fully five years after Germany’s He 176). Several more flights were conducted, some equipped with transmitters that sent flight test data to ground-based recorders: an early use of telemetry that would

MX-324 at Harper Dry Lake.

eventually enable test engineers to monitor in real time how an airplane prototype was performing.

The actual XP-79 rocket plane was never built. Delays in the development of the complicated Rotojet rocket engine, which could never be made to work at full scale, eventually led the USAAF to cancel the XP-79. The development of the jet-powered XP-79B continued. On 12 September 1945 Crosby finally took the XP-79B up from Mojave’s Rogers Dry Lake (later to become famous as Edwards Air Force Base). All was well for about 15 minutes, then the plane suddenly entered a spin from which he could not recover. Crosby failed to bail out and was killed when the plane struck the desert floor. The magnesium alloy structure was almost completely consumed by the resulting fire. After this disaster the USAAF decided to abandon the project. Flying wings like the MX-334 and the XP-79 were signature designs of Northrop and had many advantages over more conventional aircraft shapes, but they proved difficult to fly. Only when sophisticated control electronics became available did a flying wing finally become operational as the Northrop Grumman B-2 Spirit stealth bomber in 1997.

In the United Kingdom rocket aircraft developments were even more modest. A launch sled propelled by solid rocket motors was developed to ‘catapult’ especially modified Hurricane fighters dubbed ‘Hurricats’ off the bows of merchant ships as a means of protecting convoys from marauding German Condor long range bombers. Steam catapults, as used to launch small float planes from Navy ships, were too weak to get the heavy fighter airborne. The Hurricat was a stop-gap measure, as the plane could be used only once: after its mission it had nowhere to land and the pilot had to bail out or ditch his aircraft as near as possible to the convoy that he was defending, and hope that one of the ships would pick him up.


By the mid-1980s it had become clear that a spaceplane based on rocket propulsion only was very difficult to achieve, and that introducing airbreathing engines for at least part of the flight would be necessary. Whilst a single-stage orbital rocket plane must consist of at least 90% propellant, for an efficient airbreathing spaceplane this drops to less than 70%. In other words, for the same payload and propellant weight the structure weight of an airbreathing spaceplane can be more than tripled to over 30% instead of 10% of the overall weight. The dramatic relaxation of this constraint just might make a single-stage spaceplane possible, although it would require a propulsion system significantly more complex than a pure rocket design.

In Germany in 1985, Messerschmitt-Bolkow-Blohm (MBB, with which Junkers had been merged in the meantime) revived the idea for a two-stage, horizontal take­off and landing (HTHL) spaceplane. They named the new concept the ‘Sanger-IF, and its goal was to lower the launch price of satellites and other cargo by a factor of three to ten. As with the Sanger-I it would involve a large, delta-winged hypersonic carrier aircraft and a smaller orbital vehicle, but the carrier would take off from a regular runway and instead of rocket engines it would use airbreathing turboramjets with liquid hydrogen as fuel. Since no oxidizer would be carried, this would result in great mass and volume savings. However it required five very complicated engines that would work as turbojets at speeds up to Mach 3.5 and then as ramjets at higher velocities (the Lockheed SR-71 Blackbird had turboramjet engines but the Sanger-II carrier would be required to fly much faster than the SR-71’s top speed of Mach 3.2). It would be 84 meters (275 feet) long and have a wingspan of 41 meters (135 feet). After taking off from a runway in Europe the carrier aircraft would fly at Mach 4 to

Model of the Sanger-II concept [МВВ].

the appropriate latitude for its intended orbit (typically south of Europe, closer to the equator) at an altitude of 25 km (82,000 feet) in order not to pollute the critical ozone layer of the atmosphere. Next it would accelerate to Mach 7 and fly up to an altitude of 31 km (100,000 feet) to release its upper stage, which would use its own rocket engine (burning liquid oxygen and liquid hydrogen) to continue into orbit while the carrier glided back to base. In a modified form the first stage would also be able to function as a hypersonic airhner capable of flying 230 passengers at a top speed of Mach 4.4 for a distance of 11,000 km (6,500 miles); e. g. Frankfurt to Cape Town in under 3 hours. The upper stage would either be a non-reusable, unmanned ‘Cargus’ with up to 15,000 kg (33,000 pounds) of payload, or a HORUS reusable shuttle with a two-person crew and either 36 passengers or 3,000 kg (7,000 pounds) of cargo. The maximum lift-off weight of the Sanger-II would be 350,000 kg (770,000 pounds).

The German government funded a concept study in order to refine the design, as well as a technology development program that led to ground runs of Europe’s first turboramjet engine at MBB in 1991. However, it was concluded in 1994 that full development would be much too costly and the operational launch cost savings in comparison to those of expendable launchers too uncertain, so the entire project was canceled. All that seems to be left are a large scale model of the spaceplane and a laboratory-sized ramjet demonstrator, both of which are on display in the German Technik Museum Speyer (close to the jet-engined Buran shuttle).

The Sanger-II project was part of a kind of spaceplane revival that began in the mid-1980s and ended abruptly in the mid-1990s. This renewed interest was prompted by the rationale that something new was required in order to cut the costs of access to space in comparison with uneconomical expendable rockets and the Space Shuttle, and also that the technology necessary for this was now within reach. In parallel with Germany’s Sanger-II the US devoted a lot of effort to the aforementioned NASP, the British worked on their HOTOL spaceplane, and France, Japan and Russia were all independently working on reusable aircraft-like launch vehicles. It appeared that the days of the expendable launch vehicle were finally numbered.

Star-Rakers loading cargo at an airport [Rockwell International].

Even as the Space Shuttle was being developed, in the US in the 1970s there were many studies of a possible successor in the form of a single-stage-to-orbit reusable launch vehicle. The intrinsic weight issues led to the conclusion that complicated tri­propellant rocket engines would very likely have to be created for rocket propelled spaceplanes. These would initially use low-energy but high-density kerosene as fuel to generate high thrust for take-off with a limited tank volume, and then low-density but high-energy liquid hydrogen to efficiently accelerate up to orbital velocity. Sled- launch systems such as Sanger envisioned were also seen as a potential solution; for instance the Reusable Aerodynamic Space Vehicle single-stage spaceplane proposed by Boeing, which envisaged using two Space Shuttle Main Engines for propulsion. Rockwell International offered an alternative concept called the ‘Star-Raker’, a delta-winged HTHL SSTO with ten (!) “supersonic-turbofan/air-turbo-exchanger/ ramjet” engines, three large rocket motors and an undercarriage that would be jettisoned and recovered by parachute. The company issued colorful illustrations showing several Star-Rakers at a commercial airport with their hinged noses open to load cargo, thus emphasizing airline-like operations.

The various studies by NASA, the US Air Force and contracted industries in the 1970s led to a classified military program called Copper Canyon (as with most secret military programs the meaningless name was intended to mask what it was about) that ran between 1982 and 1985, and out of which the US National Aerospace Plane (NASP) emerged as the less classified follow-on announced by President Reagan in his State of the Union of 1986. NASP was to lead to an air-breathing scramjet HTHL spaceplane prototype designated the X-30, operational derivatives of which would be able to function as a single-stage-to-orbit launch vehicle, a hypersonic airliner called the Orient Express (somewhat similar to the dual use the Germans had in mind for the first stage of their Sanger-II), or a military Mach 12 reconnaissance plane and/or strategic bomber offering the response speed of a ballistic missile and the flexibility, accuracy and ‘recallability’ of a bomber. (RecallabiUty means that you can change your mind about blowing some place to bits when the bomber is already on its way, something impossible with a ballistic missile.) Early artistic impressions showed an elegant Concorde-like design which looked as sleek and fast as it was intended to become.

The program was jointly run by NASA and the Department of Defense. In 1990 Rockwell International became the prime contractor for its development. By then the spaceplane had grown considerably in weight and size in comparison to the original design of 1984. Having lost its resemblance to the Concorde, it now had a wedge-shaped aerodynamic configuration called a ‘waverider’ (essentially a hypersonic surfboard) with most of the lift being generated by a shock wave compressing the air below the plane. This shock wave, created by the forward fuselage, would also compress the air before it entered the engines, effectively

NASP as NASA imagined it in 1990 [NASA].

supplying the scramjets with more oxygen. The aft fuselage formed a gigantic integrated nozzle for expanding the scramjet’s exhaust. There were small wings to trim the aircraft and provide control. Its overall configuration was ideal for efficient high-speed flight but gave poor lift at low speeds and in particular for taking off.

Much new technology was required, including a lightweight, composite-material hydrogen tank and advanced computer programs for modeling the airflow around the aircraft and through the engines. The hydrogen fuel would require to be carried in the form of a slush (liquid and ice mixture) to limit the volume of the propellant tanks and therefore the size, and most importantly weight, of the aircraft’s structure (a kilogram of hydrogen ice has a lower volume than a kilogram of liquid hydrogen). Heat-resisting carbon materials would be needed for the aerodynamic surfaces that would endure temperatures over 1,700 degrees Celsius (3,000 degrees Fahrenheit) during hypersonic ascent and atmosphere re-entry, and titanium aluminide panels for most of the fuselage. A major hurdle was the development of the scramjet in which liquid hydrogen would be injected into the combustion chamber and be ignited by the hot compressed air rushing in at hypersonic speed. The exhaust would primarily consist of water vapor and be environmentally friendly but the decision to use hydrogen as fuel was mainly driven by the need for high performance and high efficiency.

The NASP design incorporated the clever idea of using the atmospheric heating of the vehicle to increase the thrust of its scramjet engines: by circulating hydrogen fuel through the plane’s skin to warm it up prior to injection into the engine the energy generated by atmospheric drag was effectively added to the thrust of the scramjet. At the same time the cryogenic hydrogen flow would cool the aircraft. It was initially believed that this revolutionary scramjet propulsion and temperature control system would make it possible for NASP to reach Mach 25 in the high atmosphere; enough for it to achieve orbit without the need for additional rocket engines. However, as the development progressed it became clear that the maximum speed attainable would be about Mach 17, as at higher speeds the weight of the active thermal control system would exceed that which would be required to add conventional rocket engines and propellant. Hence for NASP to be used as a single-stage-to-orbit spaceplane it would need additional rocket propulsion. In fact, rocket propulsion would be required also to get NASP up to Mach 3 for the scramjets to take over. And rocket propulsion would be required to perform the deorbit maneuver at the end of an orbital mission. To keep the vehicle sufficiently light, it was planned to use a new type of rocket engine that, apart from onboard fuel, would be fed with air liquefied during atmospheric flight by a combination of ram-compression and cooling by liquid hydrogen; an airbreathing rocket engine known as LACE (‘Liquefied Air Cycle Engine’). Updated requirements stated that the X-30 must carry a crew of two and that, although it was an experimental vehicle, it must also be able to deliver a small payload into orbit. Hence rather than a demonstrator, the X – 30 was to be a semi-operational vehicle. It was supposed to be able to fly a mission every 72 hours (compared to a maximum of one per month for the Space Shuttle) and require only 100 workers for its operations. The ensuing fully operational launch vehicle derivative was expected to cut the cost per kg of payload by a factor of ten.

Meanwhile the estimated cost of the full development into an operational vehicle had increased far beyond the originally projected $3.3 billion for a relatively modest technology demonstrator. By the early 1990s the projected cost for the demonstrator was $17 billion and the fully operational launch vehicle would require another $10 to $20 billion (in 2011 dollars those numbers would be close to 27 and 16 to 32 billion, respectively). It was also expected that another two decades of development would be required to master all the relevant technological issues prior to building a working prototype. Moreover, the argument that NASP would enable airplane-like operations that would result in low costs per flight and rapid turn-around times sounded awfully similar to those predicted but never achieved for the Space Shuttle. The use of liquid hydrogen alone would mean a complete departure from conventional airport storage and distribution facilities, essentially ruling out the use of normal airfields because it would be prohibitively expensive to equip a sufficiently large network of them with the required production, storage and handling facilities. The severe cost increase and schedule stretch, uncertain operational benefits, and the necessity to comply with too many (civil space, commercial airline and military) requirements eroded support for NASP below the critical level. In addition, the collapse of the Soviet Union severely reduced the push in the US for ambitious technical and military programs. Inevitably, it was terminated in 1993.

NASP was initially superseded by the Hypersonic Systems Technology Program (HySTP) in which NASA and the Department of Defense continued with technology development on a less ambitious scale, but when the Air Force withdrew in 1995 the development of a US spaceplane pretty much expired.

During the winter of 1993-1994 the Air Force’s Phillips Laboratory conducted a six-week study of an interesting alternative called ‘Black Horse’ which called for a single-stage rocket propelled spaceplane using hydrogen peroxide and kerosene. This non-cryogenic propellant combination is better suited than cryogenic fluids such as liquid hydrogen and liquid oxygen for rapid-reaction launches for military purposes, and furthermore has a much higher density leading to smaller tanks and therefore an overall smaller, lighter aircraft. But it lacks the specific impulse required for a single­stage spaceplane. To compensate for this lack of performance (and made practicable by the use of storable fluids) once the Black Horse was airborne it would rendezvous with a large subsonic tanker aircraft (a converted airliner) for aerial refueling prior to continuing into orbit. The spaceplane would effectively start its flight not from the ground at zero speed but from 13 km (43,000 feet) and Mach 0.85 (a more advanced system could involve a newly developed supersonic tanker able to do the refueling at a much higher speed). As a pure rocket plane the structure weight constraints for this Aerial Propellant Transfer (APT) vehicle where nevertheless severe. And although aerial refueling is standard procedure for agile military jet fighters, doing it with an

84,0 kg (185,000 pound) rocket plane would be a different matter entirely. In early 1994 Martin Marietta investigated a near-term suborbital X-plane with turbofans and liquid oxygen/kerosene rocket engines to demonstrate the APT concept. Flying up to about half orbital speed, Mach 12, this aircraft could fire an expendable upper stage to insert a satellite into low orbit, or use a Sanger-type atmospheric bounce trajectory to fly for long distances or drop bombs. Being about half the size of the Black Horse design it was called ‘Black Colt’. Several people who worked on these studies went on to establish the Pioneer Rocketplane company in 1996 to develop a commercial satellite launcher based on ATP called ‘Pathfinder’. However, they quickly found out that a switch from hydrogen peroxide to liquid oxygen as oxidizer was required to turn Black Colt into an effective launch vehicle. In the late-1990s, just as the size and complexity of the vehicle increased, the intended market of launching constellations of communications satellites evaporated due to develop­ments in less costly terrestrial mobile phone networks. Without a strong demand, the Pathfinder could no longer be pursued.

The British HOTOL (short for ‘Horizontal Take-Off and Landing’) was a concept for an unmanned, single-stage-to-orbit spaceplane able to use ordinary runways. It would require the development of a novel propulsion system that combined turbojet, ramjet and rocket engine elements. As with the Sanger-II and X-30/NASP, it would use hydrogen as fuel and draw upon atmospheric oxygen as much as possible in order to save the weight and volume of oxidizer that the vehicle would otherwise be required to carry.

In 1982 rocket engineer Alan Bond came up with the idea for a new type of engine combining airbreathing and rocket propulsion that he thought would enable a spaceplane to consist of only one single vehicle (like NASP). Around the same time Robert Parkinson of British Aerospace (BAe) was conducting a study of a reusable launch vehicle. Bond’s engine design, the rights for which had by then been bought by Rolls Royce, was combined with BAe’s launch vehicle concept and in 1985 the BAe/Rolls Royce HOTOL project was bom. It became an official national program in 1986 when the government decided to fund a 24 month proof-of-concept study led by Parkinson and John Scott of Rolls Royce.

The design that emerged looked somewhat like a torpedo with small delta wings at the rear, a single moveable fin up near the nose, and air intakes at the aft-bottom fuselage. It would be powered by four of the Rolls Royce RB545 Swallow engines, which in the atmosphere would use liquid hydrogen to pre-cool the hot air entering the engine and thereby make unusually high compression possible. The air initially entering the engine would have a temperature of about 1,000 degrees Celsius (1,800 degrees Fahrenheit) because it would arrive at high speed and then be slowed down almost to a standstill, resulting in almost all kinetic energy being converted into heat. Subsequent compression in the engine for efficient combustion and thrust generation would increase the air temperature even further, so starting off with the hot air would quickly lead to unacceptably high temperatures in the engine. However, using pre-cooling the compression would start with air which had been chilled to minus 130 degrees Celsius (minus 200 degrees Fahrenheit), yielding less extreme temperatures after high compression. In addition, cooling the air would avoid the need for heavy, high-temperature materials in the compressor section of the engine. Some of the now relatively warm hydrogen coming out of the air pre-cooler system would be used to drive the engine’s hydrogen turbopump while the rest would be burned together with the cooled air in the turbojet section of the engine at relatively low speeds and in the ramjet section at supersonic speeds. At an altitude of 26 km

The British HOTOL concept [BAe].

(85,000 feet) and flying at Mach 5 the engines would be switched to pure rocket mode to bum hydrogen with onboard liquid oxygen instead of air. Unlike NASP, there would not be any scramjet. The return flight would be completely unpowered.

From the very start, HOTOL was to be able to fly into space and back completely automatically, without a crew and with limited contact with ground control stations. Any astronaut passengers would merely be payload in a special container that would fit into the cargo bay located in the middle of the fuselage.

The development of HOTOL soon came up against a typical spaceplane issue that must have also plagued NASP and Sanger-II: the large movement of the center of gravity as the flight progressed. Most of the propellant was housed in the forward and center fuselage whilst the heavy and fixed weight of the propulsion system was all necessarily placed at the aft end of the vehicle. This meant that the plane’s center of gravity moved significantly aft as propellant was consumed. At the same time the large range of Mach numbers meant the aerodynamic center of air pressure moved significantly during the flight: it being aft while accelerating during the ascent, and forward while slowing down upon return. Another complicating factor was that the plane would take off with virtually full tanks but would later approach the airport and land with empty tanks and very little payload, so the center of gravity would thus be completely different at the same relatively low velocities at the start and the end of a mission. Balancing the centers of gravity and pressure became very tricky for the HOTOL lay-out, forcing the designers to modify the fuselage and to locate the wings far aft in order that their lift could counteract the weight of the engines and keep the plane stable at all speeds and propellant loads. But the new configuration that solved the balancing problems required a reduction of the payload. This was compensated by abandoning the conventional aircraft take-off in favor of a rocket propelled trolley (resulting in a similar take-off approach to that Eugen Sanger had in mind for his Silbervogel). This was a serious departure from the originally envisaged aircraft-like operations capability and significantly lowered the commer­cial attractiveness of the project. And when the design team nevertheless found that they had to resort to all manner of untried, experimental materials and structure technologies to maintain a decent payload at a reasonable flight price this caused the government to withdraw its support in 1988, and shortly thereafter Rolls Royce also pulled out. In an effort to save the project in 1990 BAe approached the Soviets to study launching a HOTOL-derived vehicle off the back of their Antonov An-225 (as planned for their MAKS concept). This so-called ‘Interim HOTOL’ would also abandon the complex RB545 engines and instead use conventional rocket engines, possibly of Russian pedigree. Thus HOTOL would become a pure rocket plane launched by a subsonic carrier first stage that would release it at an altitude of 9 km (30,000 feet) and a starting velocity of about Mach 0.7. Antonov also studied the possibility of fitting the An-225 with two additional jet engines (turning it into an An-325) to carry a larger version of the vehicle. But neither the UK government nor the European Space Agency expressed any serious interest and the disintegrating USSR could ill afford to participate, so in 1992 the project was canceled.

Whereas Sanger-II was to be a jet-engined hypersonic carrier aircraft for a rocket propelled upper stage, and NASP was initially conceived to be an airbreathing-only hypersonic spaceplane, the original HOTOL was to fly itself into space using rocket motors and would therefore have been a real rocket plane in that respect (albeit with additional jet propulsion). On the other hand it was to fly unmanned, making it more like a missile than a plane capable of being controlled by a pilot at will. For these spaceplanes the distinction between a jet aircraft, a pre-programmed launcher and a rocket plane becomes rather ambiguous.

The Soviets also had their own aerospaceplane projects. From the mid-1970s to the late 1980s the Myasishchev Experimental Design Bureau worked on the MG-19, a rather megalomaniac concept involving a triangular lifting body with a 500,000 kg take-off weight and a 40,000 kg payload to low orbit. It would use turbojets and then scramjets to get to Mach 16, and subsequently rocket propulsion with hydrogen fuel superheated by a nuclear reactor to achieve orbit. The complexity of the vehicle, the dangers involved in flying a nuclear reactor in a hypersonic aircraft, and the priorities of the Buran space shuttle project meant that by the 1990s this project was gone. The idea of using nuclear power in a spaceplane was not new (since the aforementioned Martin Astroplane with its nuclear magnetohydrodynamic engines had already been proposed in the US in 1961) but it is surprising that Myasishchev still considered it to be a sane solution as late as the 1980s.

The USSR’s specific response to NASP was the Tupolev Tu-2000, a long-range heavy bomber and single-stage-to-orbit vehicle whose turbojet/scramjet/rocket engines used liquid hydrogen and liquid oxygen. Development of this spaceplane started in 1986, aiming for an initial experimental two-person design called the Tu – 2000A that would be capable of reaching Mach 6. After the collapse of the Soviet Union work continued in Russia with two tests of an experimental, sub-scale scramjet at subsonic and supersonic speeds up to Mach 6 using S-200 tactical missiles in 1991 and 1992. However, in 1992 the project was suspended owing to a lack of funds. From the few concept images available it appears that the Tu-2000 would have looked very similar to NASP.

In Japan some spaceplane development was ongoing during the late 1980s and early 1990s under the designation ‘Japanese Single Stage To Orbit’ (JSSTO). Four LACE propulsion units were to drive the vehicle to Mach 5, then six scramjets would accelerate it to Mach 12. Further acceleration to orbital speed would also be achieved using the LACE engines, but now being above the ‘sensible’ atmosphere the engines would be fed air that was liquefied and stored in tanks earlier in the ascent, a process called ACE (‘Air Collection Engine’). Mitsubishi Heavy Industries tested a sub-scale scramjet in a hypersonic wind tunnel in 1994 but the work does not seem to have progressed much further than that.

In a research effort named STS (Space Transportation System) 2000, Aerospatiale in France in the late 1980s and early 1990s investigated a single-stage ramjet/rocket spaceplane that looked similar to the Concorde, as well as a Sanger-II-like concept in which a ramjet/rocket aircraft carried a rocket propelled second stage spaceplane that would be separated at Mach 6. Another French company, Dassault Aviation, worked on a Mach 7 scramjet aircraft that would air-launch an expendable Ariane 5 second stage carrying a Hermes-derived shuttle. This was called STAR-H, for ‘Systeme de Transport spatial Aerobie Reutilisable – Horizontal’ (i. e. airbreathing reusable space transportation system for horizontal take-off and landing). These French concepts were not particularly new, all resembling (aforementioned) ideas from the mid-1960s such as the European Space Transporter, the Dassault Aerospace Transporter and the various two-staged spaceplanes studied by the British Aircraft Corporation.

Of all the spaceplane projects of the 1980s, NASP, a single-stage-to-orbit vehicle with (at least initially) only airbreathing propulsion, appears to have been the most ambitious and most complex. Nevertheless, despite the large amount of technology development that was carried out, not much of this is in the public record; the project was so advanced that even today many of its details remain secret. The Sanger-II was theoretically the least complex by using two stages instead of the more constraining single-stage-to-orbit approach and incorporating airbreathing and rocket propulsion in two separate vehicles. Sanger-II’s first stage would not fly fast enough to require exotic thermal protection materials and scramjets; conventional titanium panels (with additional carbon layers at the hot-spots) and relatively simple ramjets would suffice. During re-entry, the vehicle, being mostly large and now empty tanks, would have a significantly lower density than the Space Shuttle Orbiter (hence a larger surface area in comparison to its weight) and so would encounter relatively benign temperatures. In contrast to NASP it would have been able to put substantial payloads into orbit and operate as a real launch vehicle, potentially beating everybody else to the market. However in the end all of the spaceplane concepts proved to be too far ahead of their time: scramjet and combined propulsion technology was not yet sufficiently developed and the computer programs for simulating airflow and combustion at high speeds and temperatures were not mature enough. An article on NASP in the magazine Flight International in October 1987 had already included an ominous warning: “The odd thing is that the excitement is based almost entirely on theoretical research and small-scale laboratory work. Nobody has run a Mach 25 scramjet continuously for more than a few seconds, and no powered atmospheric vehicle has attained anything like the speeds envisioned for NASP.” Nevertheless, in March 1992 Popular Mechanics enthused “Space Race 2000 is on”, anticipating one of the international spaceplane contenders developing a real vehicle by the turn of the century. But by the mid-1990s all the spaceplane projects had more or less gone: the ideal, airliner-like spaceplane as described in the Introduction of this book was (and still is) a long way off.

Part of the issue is that whereas for a conventional aircraft the design can be split into elements like fuselage, wings and engines, and each can (to some extent) follow an independent development track, a spaceplane requires a fully integrated approach in which any small change of one item can have dramatic implications for the rest of the design. For HOTOL for instance, it turned out that for each additional kilogram of inert weight that the vehicle design gained (say, an additional piece of electronics or thermal insulation) some 25 kg of additional propellant would be needed to restore the vehicle’s performance; and additional tank and structure mass would be required to house this extra propellant, which itself further increased the overall weight. It was a vicious circle. Another parameter with a dangerously strong growth factor was the specific impulse of the rocket propulsion, where a change of just 1 % would impose a 4% change in the gross lift-off weight. Therefore a slight increase in rocket engine performance would pay off but over-optimism concerning the engine’s performance could make the vehicle considerably heavier, larger and more expensive. Accepting a slight reduction in speed to compensate for weight growth is possible in the design of normal aircraft, but a no-go for a spaceplane that needs to achieve a fixed minimum velocity if it is to enter orbit. As only 3% of HOTOL’s take-off weight would consist of payload, it would not require many design changes to consume all of this weight allocation for useful cargo and thus render the vehicle completely useless. The Concorde had a 5% payload but its design was considerably less sensitive to changes because it flew much slower than HOTOL and had more scope to trade-off between payload, speed and range; flying slightly slower than originally envisaged does not a ruin an aircraft design, but for a spaceplane it means the difference between achieving orbit or not.

The very high sensitivity of its performance to the input assumptions makes the design of a single-stage-to-orbit vehicle particularly difficult, and obliges engineers to pay very close attention to the fine detail. It also requires that a wide range of new technologies be advanced to maturity before the spaceplane design can be finalized, because any small discrepancy in expected weight or performance can have major consequences for the entire vehicle design. All this, plus the need to make the vehicle easy to maintain and operate, makes clear just how daunting the task of developing a spaceplane is.

The rise and fall of the rocket interceptor

“The only time you have too much fuel is when you’re on fire.” – Anonymous

The world had entered the Second World War with piston-engine propeller aircraft technology, and these were sufficient for the duration of the war. However, shortly after Germany’s surrender the former Allies suddenly found themselves in the jet age, as well as kicking off a conflict between the US and Western Europe on the one hand and the Soviet Union and Eastern Europe on the other. It was clear that any new air war would be fought by planes equipped with turbojets and/or rocket engines: no piston-engine aircraft could ever hope to keep up with the new jet fighters that were already flying at the end of the war, particularly the German Me 262 and Domier 234 and the British Gloster Meteor turbojet.

Because of this paradigm shift in aircraft propulsion, advanced but conventional propeller fighters developed at the end of the war became obsolete almost overnight, just as a new generation of aircraft became necessary in order to maintain the uneven military balance of the Cold War and to fight in the limited-scope conflicts that this spawned (such as in Korea and the Middle East). It was soon recognized that major development efforts were required for the new generation of high-speed fighters, but whether these would primarily be propelled by rocket or jet engines remained to be determined.

The need for rapid interceptors quickly became very urgent in Europe because of the proximity of the countries of the Warsaw Pact and NATO: their bombers could attack one another’s cities and military facihties within minutes of crossing the Iron Curtain. Moreover, an incoming bomber could spend a lot of time getting up to high altitude and speed in friendly airspace prior to crossing the border, but a defending interceptor would have little time to react. In fact, the situation was very similar to that which had faced the German Me 163 pilots in 1944 and 1945. Interceptors which could achieve high speeds and high altitudes in little time were therefore a priority in post-war Europe.

As we have already seen, the push for fast developments in aeronautics during the 1940s meant that new aircraft designs followed one another in rapid succession. This situation would continue into the 1950s and 1960s, with concepts sometimes already being obsolete prior to their first flight. Amidst this design fury, rocket aircraft had to compete with turbojet aircraft for development funding.

By 1945 several rocket planes had been successfully flown, starting with simple gliders fitted with solid rocket boosters, through modified piston-powered planes to dedicated rocket propelled aircraft. However, only the Me 163B had seen operational service. Overall, this experience did not bode well for the rocket powered interceptor concept. There were many accidents due to the poorly understood aerodynamics of transonic flight. Also the propellants tended to be difficult to handle and downright dangerous (notably the corrosive hydrogen peroxide of the German Walter engines and the almost equally nasty nitric acid used by the Russian engines), especially in combination with the rather low reliability of the rocket engines used. Moreover, the very short range of the rocket aircraft restricted it to quick attacks on enemy aircraft flying in the vicinity of the interceptor’s base. Protecting an entire country such as Russia would require vast numbers of rocket interceptors and airfields. Furthermore, Germany’s investments in often rather fanciful ‘wonder aircraft’ and rockets were of little help during the war: if the Luftwaffe and the Wehrmacht (Army) had bought conventional weapons for the money they invested in highly novel technology, they might have been able to extend the war considerably. In other words, the value of rocket aircraft remained to be proven. However, near the end of the war the problem of how to efficiently operate rocket airplanes did not seem to be so much intrinsic to the type of technology than to its immaturity.

The aerodynamic problems associated with transonic speeds, which were an issue for new jet aircraft as well, could only be solved using the proper wind tunnel tests, theoretical modeling and innovative wing design. Of course, the route to supersonic flight had already been partly explored by Lippisch and other aerodynamicists. The German, Japanese and Russian high-speed interceptors of that time looked distinctly futuristic, but their swept-back or delta wings, tail exhaust nozzles, cockpits placed in front of the wings and lack of propellers would soon become the norm for fighter aircraft. Some of the early jets and rocket aircraft still look rather modern today! As regards the unpleasant rocket propellants, fortunately less vicious alternatives were able to be produced.

The only true disadvantage inherent in rocket power was short range, limiting the role of rocket-propelled fighter planes to point defense: an interceptor used to defend a specific target by taking off and climbing to altitude as rapidly as possible in order to counter an approaching threat, and then land and prepare for another mission. But it had already become apparent during the war that the range issue could be partly resolved by a combination of rocket and jet engines, with the relatively simple rocket providing the thrust for a rapid ascent and the more complicated jet engine enabling the plane to cruise for relatively long periods.

Because of the rapid developments during the war, jet engines had become more competitive, in terms of thrust per engine weight, in comparison to rocket engines. The early Heinkel HeS 3 jet engine that powered the world’s first jet aircraft in 1939, the He 178, weighed 420 kg (930 pounds) and produced a maximum thrust of 4,400 Newton (equivalent to about 440 kg). This meager thrust to weight ratio meant the

engine could vertically lift just 1.05 times its own weight. In 1941 the ‘cold’ Walter RII-203 rocket of the Me 163A was far better. It had a thrust of 7,400 Newton and weighed only 76 kg (170 pounds), which is a thrust to weight ratio of around 10. The early rocket engine could thus produce some ten times more thrust per unit of engine weight than the early jet engine and was simpler to manufacture and maintain. The jet engine was however more fuel-efficient, requiring less propellant than the rocket engine for the same amount of thrust. Near the end of the war the Junkers Jumo 004D turbojet engine, weighing 745 kg (1,640 pounds), could already deliver a thrust of 10,300 Newton at low altitudes and had a thrust to weight ratio of about 1.4. The Me 163B’s Walter HWK 109-509A2 of 1944 still had a higher maximum thrust of

17,0 Newton, weighed only 160 kg (350 pounds) and so had a thrust to weight ratio of about 11. Then again, at least in terms of thrust, the difference was shrinking. Not only were jets more economical in propellant consumption than aircraft rocket engines (translating into longer flights and greater ranges), they were now producing similar amounts of thrust while their thrust to weight ratio was rapidly improving.

A disadvantage early jet engines had with respect to rockets was that they required very careful throttle movements: changing power rapidly would often extinguish the combustion in the turbojets (so-called flameouts). The start procedures for early jet engines were also very sensitive, as there was no automatic sequencing of the various events built in. An error could cause the engine to quit or possibly overheat. Even worse for the pilots was that everyone on the airfield would know about his mistake because the engine would produce a loud rumble and shoot a huge flame out of the tailpipe. Rocket engines were in general much less fidgety. However, the sensitivity issues of turbojets were soon solved by more advanced fuel controls.

In other words many advantages rocket engines still had over jet engines for use in fighter planes were being eliminated one after another. One strong inherent plus was the lack of the need for an inflow of air and associated air intakes, which gave rocket aircraft an advantage in terms of aerodynamic drag. Air intakes also cause technical complexity because the air going into a conventional jet turbine engine needs to flow at subsonic speed even when the plane itself is flying supersonically. Hence using rocket engines also meant that the propulsion was independent of the vehicle’s flight speed. A more important remaining advantage of rocket engines was that they were much more effective than jet engines at high altitudes. Rockets have their own oxidizer whereas jets need to scoop up air to use its oxygen. The density of the atmosphere drops with altitude, so for each jet engine there is an altitude at which it can no longer be fed sufficient air to work properly. Rocket engines can even work in vacuum, and in principle actually work better there since their exhaust flow is no longer hindered by air blocking the nozzle’s exit. As a result, the thrust of a jet engine decreases with altitude whilst that of a rocket engine increases (by up to 25%, at least for a rocket engine with a nozzle that is optimized for use at high altitudes).


Although the large spaceplane programs were canceled, some related developments did survive. Work on hypersonic scramjet propulsion in the US was continued in NASA’s Hyper-X program, and resulted in two test flights of the small X-43A. This unmanned experimental vehicle was launched from the nose of a Pegasus rocket that was itself dropped from a converted airliner, and set new records for an airbreathing vehicle by achieving Mach 9.6 and a scramjet burn of 10 seconds on its latest flight in November 2004. Scramjet development tests are continuing with the similar X-51, which flew for the first time in May 2010 and reached a speed of ‘only’ Mach 5 but a much longer powered-flight time of 200 seconds. The second flight, in June 2011, was unsuccessful due to a failure of the scramjet engine. Some people insist that the US military already has a highly secret orbital spaceplane in operation (such as the ‘Blackstar’ reported by Aviation Week) but the evidence is unconvincing.

Shortly after HOTOL foundered in 1988, members of the engine design team led by Alan Bond set up a new company (Reaction Engines Ltd) to continue to develop the HOTOL concept, focusing initially on an improved version of the RB545 engine called the SABRE (‘Synergistic Air Breathing Engine’) and in particular the crucial precooler section. The spaceplane concept currently being worked on is the Skylon presented in the Introduction of this book as a “perfect spaceplane” (it is named after a futuristic art structure included in the 1951 Festival of Britain, which the fuselage strongly resembles). The designers reckon they have fixed the flaws in the HOTOL design, in particular the stability problem due to the heavy engines in the aft part of spaceplane. The Skylon solution is to locate the engines in the middle of the vehicle, housing them in nacelles at the tips of the delta wings in the same way envisaged for the Keldysh Bomber in 1947. This prevents the center of gravity from moving aft as the propellant tanks are depleted. Moreover, since the engines are not fully integrated with the fuselage they can be tested separately from the remainder of the vehicle. The engine nacelles have a peculiar banana-shape because their air intakes have to point directly into the airflow, whereas the spaceplane’s wings and body must fly at an angle to create lift. Each engine will give a maximum thrust of

1,350,0 Newton in airbreathing mode, and 1,800,000 Newton in rocket mode.

According to the company their SABRE propulsion would make Skylon very safe and reliable, and enable it to take off without the rocket trolley that would have been necessary for HOTOL. But this meant Skylon would need a sturdy undercarriage as well as strong brakes to stop itself before the end of the runway if a problem were to occur just before take-off. It was decided to cool the brakes by water, which would boil away and dissipate the heat caused by the braking friction. The cooling water would be jettisoned following a successful take-off, thus reducing the weight of the undercarriage by several metric tons. At landing Skylon would be empty and hence fairly Ught, so the brakes would not need water cooling in order to be able to stop the plane without catching fire. Due to its aerodynamic characteristics upon re-entry, the vehicle would slow down at higher altitudes than the Space Shuttle Orbiter, keeping the skin of the vehicle significantly cooler, hence requiring only a durable reinforced ceramic for most of its skin. The turbulent airflow around the wings during re-entry

Model of the SABRE engine [Reaction Engines Limited],

Oxygen Payload Sabre Tank Container Engine

Cutaway of the Skylon concept [Adrian Mann & Reaction Engines Limited],

would, however, necessitate active cooling of some parts of the vehicle. Skylon is expected to be able to put 12,000 kg of payload into low orbit. Its take-off noise is expected to be acceptable for taking off from regular airports in populated areas but the runways would have to be extended to 5.6 km (3.5 miles) in length, of which the first 4 km (2.5 miles) would require to be stronger than usual to cope with the heavily laden Skylon rolling at high speed.

An independent review by the European Space Agency, which is also funding part of the technology development for Skylon, concluded in 2011 that the overall design “does not demonstrate any areas of implausibility”. Reaction Engines is confident that Skylon will soon reach a technical maturity sufficient to convince investors that it is a valid commercial opportunity which warrants funding to full development. The project’s cost estimates indicate that if a fleet of 90 vehicles were produced it would be possible to buy a Skylon for about $650 million, which is roughly comparable to a large jet airliner. Early customers would pay $30 to $40 million per flight but with more aircraft flying and an increasing total number of flights, the price should fall to around $10 milhon per launch. In comparison, a current Ariane 5 expendable rocket costs around $150 million per launch and puts less mass into a low orbit.

At the time of writing, Reaction Engines is doing tests on an experimental version of the precooler and plans to build a sub-scale version of the SABRE to demonstrate (on the ground) the complete engine’s airbreathing and rocket modes as well as the transition between these. Tests of the nacelle in which the SABRE is to be housed are to be performed using a Nacelle Test Vehicle (NTV). This is planned to be launched from the ground and use rocket engines to get up to Mach 5, at which speed internal ramjet combustion systems will simulate the operation of the air-breathing engine. The remainder of the nacelle test article will be as close as possible to the real thing, including the control systems and internal flow ducts. The NTV is also to get some data on shock interactions between the nacelle and the Skylon’s wing. According to Bond, “we could have a Skylon plane leaving Heathrow airport sometime during this century”.

In 1991 Bristol Spaceplanes, another small aerospace company in the UK, began working on rocket plane concepts. Their ‘Spacecab’ design involves a Concorde-like carrier aircraft that uses four ordinary turbojet engines to take off and accelerate to Mach 2 and two rocket engines to reach Mach 4. At that speed a small, delta-winged rocket propelled orbiter carrying two pilots and either six passengers or cargo would separate and climb into orbit. The company insists this is a very conservative design that does not require any new technology to be developed. A next-generation vehicle called ‘Spacebus’ would fit the carrier aircraft with newly developed turbo-ramjets to achieve Mach 4 and rocket engines for Mach 6. Its enlarged orbiter would have room for fifty passengers. David Ashford, the company’s managing director, has published his ideas in two popular science books: Your Spaceflight Manual – How You Could be a Tourist in Space Within Twenty Years (written with space tourism promoter Patrick Collins) and Spaceflight Revolution.

Private companies like Reaction Engines and Bristol Spaceplanes keep working on spaceplane technology, and the development of hypersonic and scramjet engines for military applications is strongly supported in the US (specifically in the Falcon program). Non-military government space agencies such as NASA and the European Space Agency have not completely given up on reusable launch vehicles either. The Future European Space Transportation Investigation Programme (FESTIP) study by ESA in 1994-1998 for instance, investigated many basic reusable launcher concepts, several of which were spaceplanes. In 2005 ESA also ran a small internal study (in which I participated as System Engineer) for a small rocket plane called Socrates. This was intended to fly at speeds up to Mach 12 and be operated for about 30 flights. It was specifically to investigate spaceplane operations and maintenance, such as how long it takes to replace rocket engines, how an onboard health monitoring system could help speed up maintenance, and how long thermal protection materials could last. A reusable rocket plane’s ability to make many test flights at relatively low cost should result in a higher reliability in comparison to modern expendable launchers which are usually deemed operational after only a single qualification flight but typically have a failure rate of 2 to 3% (meaning that two or three payloads per hundred are lost, in turn pushing up insurance costs).

India, a country that is making great strides in spacecraft and launcher technology, is investigating a concept known as AVATAR (for ‘Aerobic Vehicle for hypersonic Aerospace TrAnspoRtation’). This vehicle would take off using airbreathing turbo-ramjet engines and full tanks of fuel but none of the liquid oxygen that it would later need for its rocket propelled flight phase. Instead, during atmospheric flight separate ram air intakes collect air that is subsequently liquefied using liquid hydrogen-cooled heat exchangers; similar to the ACE principle of the JSSTO spaceplane concept. But unlike JSSTO, AVATAR involves another step in which the liquid oxygen in the air is mechanically extracted and stored in the previously empty oxidizer tanks so that they will be full by the time AVATAR requires to switch over to rocket propulsion. Although this is an extremely ambitious project, India is developing its capabilities at a rapid pace.


At the end of the war Russia captured from Germany three Me 163B interceptors and seven Me 163S glider trainers. These were tested in the large TsAGI wind tunnels as well as in gliding flight but a lack of the necessary propellants and unfamiliarity with the Walter engine prevented powered flights. The TsAGI analyses confirmed that the aerodynamics of the Me 163 were very suitable for high-speed flight. The Russians were of course even more interested in the Komet’s advanced successor, the Me 263, having captured some equipment and documentation along with technical staff who had worked on the project. The brewing Cold War and the threat of Western bombers renewed Soviet interest in fast, high-altitude rocket interceptors. Already in February 1946 the Lavochkin OKB was ordered to develop such an interceptor to detect and destroy enemy bombers. It had to be able to reach an altitude of 18 km (59,000 feet), achieve a top speed of 1,100 km per hour (680 miles per hour) in level flight at 5 km (16,000 feet), and be able fly at maximum thrust for six minutes. The plane was also to be capable of attacking at any time of the day and in all weather conditions, which meant that it had to be equipped with radar. Its armament was to be six 83-mm TRS-82 rockets. And to make the challenge even more difficult, the military wanted flight testing to start on 1 May 1947.

By the end of 1946 Lavochkin had finished the design for ‘Aircraft 162’: an all­metal plane powered by the familiar Dushkin RD-2M-3V, with a pressurized cockpit and air-to-air rocket launching tubes integrated in the fuselage. It was calculated that the rocket thrust would enable the plane to reach an altitude of 18 km (59,000 feet) in 2.5 minutes. Inspired by captured German research, the wings of the Lavochkin 162 were extremely innovative: not only were they swept, they were swept forward. Such wings provide most of the favorable high-speed characteristics of swept-back wings and offer the benefit of making the aircraft less sensitive to stall. When a plane with swept-back wings stalls (i. e. when the air can no longer correctly follow the contours of the wing because the angle of attack is too high) the air usually starts to let go at the root of the wing, and this effect quickly travels along the length of the wing to its tip causing the whole wing to rapidly lose lift and making the plane uncontrollable. When a forward-swept wing stalls, the disruption at the root cannot easily progress to the wingtips and ailerons. Giving a pilot much better control at high angles of attack makes his plane more maneuverable, which is very welcome on a fighter aircraft. But forward-swept wings were extremely difficult to build using standard 1940s aircraft materials since they require to be extraordinarily stiff. Otherwise, when the wingtips bend up or down at high speeds the airflow can rip the wings off (on a swept-back wing the airflow tends to push a twisting wingtip back to the horizontal). Because of this, along with the many unknown aerodynamic characteristics of swept wings, the La-162 was soon redesigned with well-understood straight wings.

A full-scale mockup was built but the chief designer, Semyon Lavochkin himself, soon began to express doubts about pursuing the concept. The plane would have a phenomenal rate of climb but its flying time and range would be extremely short. He believed further development of jet engines would result in better interceptor fighter designs. He was also concerned about the nitric acid propellant, which had proved to be very corrosive and unsafe in experiments during the war with RD-1 engines fitted in La-7 and La-120 aircraft. And there was the issue of the high operating pressures in the various pipes in the engine, which made it prone to leaks that resulted in a low reliability and heavy maintenance. Another issue was that the turbopump of the RD – 2M-3V would have not only to feed propellants to the combustion chamber but also drive an electric generator, which was calculated to be too weak to provide sufficient electrical power for the onboard radar. Furthermore, it was soon found that the high aerodynamic drag of the new straight wings would prevent the rocket engine from driving the plane to the required transonic flight speeds. All this, in combination with the troublesome development of the designated radar system, led Lavochkin to call a halt to work on the 162.

However, there was another Russian design bureau still working on a pure rocket – powered interceptor. The Mikoyan and Gurevich (MiG) design bureau had set out to develop a local version of the German Me 263 rocket fighter called the MiG 1-270. The Soviet military pushed their aircraft designers to copy as many as possible of the advanced German design features to get a head start in the new jet age. But Russian rocket engines were based on very different propellants than the Walter engine, and designers strongly resisted a radical switch from their own known and proven rocket technology. In addition, the MiG engineers and TsAGI aerodynamicists were not yet completely familiar with the characteristics and peculiarities of swept and delta­shaped wings and therefore preferred to employ conventional straight ones (the same was done in the US for the X-l experimental rocket plane, for the same reason).

Planform of the MiG 1-270.

MiG 1-270.

As a result the 1-270 essentially became an Me 263 with a Russian engine, a more conventional wing and tail design and a longer fuselage. The tail was different in that it had a horizontal stabilizer high on the vertical fin in a T-arrangement to diminish aerodynamical interference from the main wings. But the new plane had an ejection seat, a first for Soviet fighter aircraft and essential for bailing out at transonic speeds. In this respect it was better than its German predecessors. The rocket engine was the RD-2M-3V (described earlier) with two combustion chambers. It gave the 1-270 less thrust than the original HWK 109-509C of the Me 263. While the main combustion chamber of the German engine provided 20,000 Newton its counterpart on the RD – 2M-3V could yield only 11,000 Newton. The smaller ‘cruise flight’ chambers of the two engines were comparable at about 4,000 Newton. Even so, the combined thrust of both RD-2M-3V chambers was still expected to give the plane a maximum speed close to supersonic and make it possible to reach an altitude of 17 km (56,000 feet) in 3.2 minutes. For ease of maintenance and replacement of the engine, the fuselage could be split into two sections. Power came from a generator connected to a small propeller on the nose (like in the Me 163B) in addition to a generator cleverly connected to the turbopump of the rocket engine. The cockpit was pressurized, and armor plate in the forward structure and an armored windscreen would protect the pilot from enemy gun fire. Like the Me 263, the MiG 1-270 had a tricycle undercarriage. The two main wheels retracted into the fuselage since there was no room for them in the thin wings. As a result these wheels had a very narrow separation, making the plane wobbly on the ground and difficult to land on a rough field or in the presence of a cross wind. The maximum take-off weight was 4,100 kg (9,100 pounds), of which 2,100 kg (4,700 pounds) was propellant. It was armed with two 23-mm NS-23 cannon and eight RS-82 solid propellant air-to-air rockets, and it was meant to defend large industrial installations as well as military bases. The threat was anticipated to be American B-29/B-50 Superfortress bombers and the new B-36 Peacemaker that could fly at altitudes up to 15 km (48,000 feet).

Two prototypes were built, Zh-1 and Zh-2. The first was intended for gliding tests and was fitted with a ballast mass instead of an engine. Pilots trained for the gliding flights on a Yak-9 fighter used as a glider, loaded with lead weights to reproduce the weight and balance of the 1-270. Both the Yak-9 trainer and the glider version of the 1-270 were towed into the air by a Tu-2 bomber. The first glide test of the 1-270 Zh-1 occurred on 3 February 1947 with test pilot V. N Yuganov at the controls. Until July of that year the Zh-1 remained connected to the tow airplane for the entire flight, but for the second phase of the unpowered tests it was released at altitudes of 5 to 7 km (3 to 4.5 miles) to ghde home and land. It reached a maximum speed of 600 km per hour (370 miles per hour) during these unpowered tests. In early 1947 the Air Force put pressure on MiG’s chief designer, Artem Mikoyan, to get the 1-270 ready for a powered flight demonstration during the annual air display in Moscow on 18 August. The gliding flight test phase was curtailed and an RD-2M-3Y engine installed in the second prototype, Zh-2, in May. Unfortunately the smaller combustion chamber blew up in ground tests on 16 July and damaged the tail section. The aircraft was repaired but could not be readied in time for the air display. An aircraft exploding during that international showcase would not make a good impression on the Soviet leadership and the rest of the word!

On 26 August, well after the show, the Zh-2 made its first two powered taxi runs and a short hop with A. K. Pakhomov at the controls. Its first (and unfortunately also last) flight was on 2 September and it lasted just 7 minutes. Pakhomov successfully climbed to an altitude of 3 km (10,000 feet), then initiated a gliding descent back but widely overshot the planned landing point and crashed beyond the airfield perimeter. He was unhurt but the aircraft was damaged beyond repair. Fortunately by then the first prototype had been upgraded by replacing the dummy engine with a real one, and powered tests resumed using the Zh-1 with V. N. Yuganov at the controls. A taxi run was made on 29 September. On the first flight on 4 October it took off under the roaring combined thrust of the two combustion chambers. The main chamber was shut down 130.5 seconds into the flight at an altitude of 4,450 meters (14,600 feet). The plane reached a maximum speed of 615 km per hour (382 miles per hour) at an altitude of 2,900 meters (9,510 feet) operating on the smaller combustion chamber only. As he glided back Yuganov found that the undercarriage would not extend so he made a belly landing that was so soft that the plane was barely damaged. The poor reliability of the RD-2M-3V rocket engine however continued to slow down the test flight phase when its largest combustion chamber exploded during a ground test and blew off a large part of the tail. The Zh-1 was repaired but was not ready for its third powered flight until January 1948. By then, however, a new problem had shown up: after each flight the engine had to be rinsed with water to remove the dangerous and corrosive nitric acid propellant, but during the freezing Russian winter this was impossible without also turning the motor into a block of ice. There was no alternative to postponing further testing until March.

In the meantime, the Aviation Technology Committee of the Air Force carried out a reassessment of the 1-270’s potential as an operational interceptor, with shatteringly negative conclusions. A major complaint was that the engine couldn’t be restarted in flight without the high risk of an explosion caused by nitric acid accumulating in the combustion chamber between extinction and reignition. The corrosive and dangerous nature of the propellant was another major issue: the I – 270’s parts and particularly its oxidizer tank, as well as the technicians working on the plane, were constantly under attack by corrosive nitric acid vapors. Personnel dealing with the engine had to wear bulky protective suits that made work on the plane arduous. The acid tank required labor-intensive removal, checks and replacement every 2 months. Special materials and acid-resistant coatings (as many as nine layers for the most critical areas) were used but corrosion occurred anyway. The difficulty in landing an unpowered 1-270 (as shown during its first powered flight) and the impossibility of rinsing the engine with water during the winter were also Usted as severe operational limitations. It is clear that the Air Force had lost its enthusiasm for the design.

By then the idea of an interceptor powered only by a rocket engine was rapidly becoming outdated. The turbojet-propelled MiG 15 had already flown in December 1947, would soon enter service, and would satisfy most of the requirements initially set for the 1-270. The MiG 15bis version had a nearly supersonic maximum speed of 1,075 km per hour (668 miles per hour) and a flight ceiling of 15.5 km (50,900 feet). Furthermore, it had a maximum range of 1,200 km (750 miles), which could even be extended to 1,980 km (1,230 miles) using externally carried drop tanks. That made it much more useful than the 1-270, which was basically only capable of a brief sortie to attack enemy aircraft that came within several tens of kilometers of its base. The much longer powered-flight endurance of the jet also meant it could attack bombers multiple times. It would even have time to dogfight with enemy fighters, something that was becoming an important requirement. Whereas in powered flight the earlier Me 163B could easily leave behind the enemy propeller fighters that it encountered, the 1-270 would have had to engage new Western jet fighters which could match it in horizontal flight. The only advantage the MiG 1-270 still had over the MiG 15 and other early jets was its rate of climb. The MiG 15 could reach an altitude of 15 km (49,000 feet) in about 5 minutes, but the 1-270 could get there in 3 minutes (with its more powerful engine the Me 263 would have done it in 2 minutes). But this single advantage didn’t outweigh all the weak points of the 1-270 and the strong points of the MiG 15.

Moreover, the development of surface-to-air guided missiles was advancing at a great rate. The S-25 (NATO codename SA-1 Guild) and the infamous S-75 (SA-2 Guideline) entered service during the 1950s and could reach and destroy high-flying intruders even more rapidly than a rocket propelled interceptor because they had no pilot for whose survival acceleration levels had to be kept to a reasonable maximum. Between them the MiG 15 turbojet and the S-25 and S-75 missiles rendered the 1-270 obsolete, with the inevitable abandonment of the idea of a manned rocket interceptor. The MiG design bureau focused its efforts on jet interceptors and fighters as the new generation of military aircraft.

The pure rocket fighter was briefly revived in the late 1940s at the OKB-2 design bureau, where a team led by A. Ya. Bereznyak (designer of the wartime BI rocket plane) and the German Siegfried Gunther jointly worked on a multi-chamber rocket interceptor. During the war Gunther had been a leading designer at the Heinkel company, and after failing to find a job in the US or the UK had offered his services to the Soviets. Gunther and Bereznyak worked on different but similar versions of their supersonic rocket aircraft indicated by project numbers 466,468 and 470. These were all delta-winged designs powered by a rocket engine with four combustion chambers installed in the tail. The maximum total thrust of this engine was to be

82,0 Newton at sea level increasing to 96,000 Newton at an altitude of 20 km (66,000 feet). The plane had no horizontal stabilizers (typical of delta-winged aircraft) but had an enormous central vertical fin as well as two smaller fins under the wings. It would take off with a jettisonable dolly or a sled driven by solid propellant boosters, and land on skids. The pilot would have a pressurized, armored cockpit equipped with an ejection seat. A radar in the nose would help him to find his target, which he would engage using either four 23-mm cannon or two canisters with six unguided missiles each. A strike version of the aircraft carrying four high – explosive bombs was also considered as an option. The 470 interceptor was envisaged to fly at a top speed of 1,910 km per hour (1,190 miles per hour) above 11.5 km (38,000 feet) altitude, equivalent to Mach 1.8, and to climb to 20 km (66,000 feet) in 2 minutes and 14 seconds. But none of these aircraft designs made it off the drawing board. Inevitable competition with surface-to-air missiles and jet fighter designs promising much larger range meant that the entire project was judged to be obsolete by June 1951. OKB-2 was disbanded and its personnel transferred to other activities. Gunter returned to West Germany in 1954 and joined Heinkel AG when that company reopened in 1955. He never spoke about his work in the USSR, and neither did the Russians, but it seems that he had worked on various other advanced Soviet aircraft apart from the 466/468/470. In 1950 Ernst Heinkel said of his former employee, “I am convinced that Giinter worked on those Soviet designs that today have become a problem for the Western World.”

The British also initiated a military rocket aircraft project just after the Second World War. The Fairey company planned a new delta-winged research aircraft called the Delta 1, which they initially envisaged as taking off vertically from a very short and steep ramp. It was to lead to an operational interceptor that would use a similar mode of take-off. The company was granted a contract in July 1946 to develop an unmanned rocket aircraft to test the concept. Simply called the Fairey VTO (Vertical Take-Off) it had an Armstrong Siddeley Beta rocket engine which was derived from the Walter HWK 109-509 ‘hot’ engine and used the same propellants. However, the Beta had two combustion chambers and nozzles each developing a thrust of 4,000 Newton, and these could be independently swiveled, one side-to-side and the other vertically so that, together, they could steer the aircraft (by what would now be called thrust vector control) at low velocities. This was important, because when the VTO left the ramp it would not be flying fast enough for its aerodynamic control surfaces to have sufficient ‘grip’. Two 3,000 Newton solid propellant boosters were added for take-off. The starting method and propulsion were so similar to the German wartime Natter that the Fairey VTO soon gained the nickname ‘Son of Natter’.

Tests started in 1949 from a ship moored in Cardigan Bay in Wales, then in 1951 resumed at the vast Woomera Rocket Range in Australia. There were stabilization

The unmanned Fairey VTO [Fairey Aviation Company].

problems early on because the autopilot, which was derived from that of the German A4/V2 rocket, had to be carefully adjusted to accommodate the completely different aerodynamics. When this was finished the take-off launches and ensuing flights were satisfactory. About 40 of these expendable test vehicles were built and launched, but it was eventually decided that the manned Fairey Delta 1 should be a jet-powered research vehicle that would take off in a conventional manner from a runway. As in the USSR, unmanned surface-to-air missiles had already taken over the role intended for the VTO rocket interceptor.


In spite of the multiplicity of studies and technology developments, government and industry funding for spaceplanes and other types of reusable launch vehicles remains modest at best: space agencies and the launch vehicle companies always seem to opt for a conventional, expendable rocket as their next generation launcher; spaceplanes are perpetually the next-next generation, with the result that for the last 50 years their full development initiation has always been 20 years in the future. This isn’t due to a lack of concepts, because in addition to the spaceplane proposals described in this chapter there are literally hundreds of ideas and designs at a variety of levels of maturity and realism. David Ashford of Bristol Spaceplanes reckons that the lack of progress on orbital spaceplanes can be attributed to an entrenched mind-set of the world’s space agencies and the vested interests of launcher industries, leading them to continue to pursue improvements and cost-reductions for their expendable launch vehicles rather than to replace these with something better. But it seems to me that really the same issues that killed the high-profile spaceplane concepts of the 1980s are still the root of the problem: uncertain economic benefits, very high development costs and great technical and financial risk; a lethal mix for any project that is not driven by a strong military or political agenda such as the Manhattan atomic bomb development of the 1940s and the Apollo lunar program of the 1960s respectively.

Reusable launchers are more expensive to develop than expendable ones, since in addition to the difficulty of developing something that can go into orbit a spaceplane must also be designed to come back, which involves re-entry into the atmosphere, a descent phase and a soft landing. Furthermore, using such vehicles requires a large infrastructure involving a long runway, facilities for vehicle and engine maintenance, (cryogenic) propellant production factories and storage tanks, and logistical systems to manage the distribution of spare parts. Large aircraft like the Airbus 380 and large launchers like Ariane 5 typically cost $10 billion to develop and it is hard to imagine how a spaceplane that combines the functions of both these vehicles is going to cost less. In fact, it is easier to see how it would cost significantly more. For instance, the reusable Venture Star SSTO abandoned in 2001 due to the expensive problems with its X-33 precursor was expected to cost close to $35 billion to make operational, and its price would certainly have increased if the additional developments to overcome the X-33 problems were carried out.

As regards the recurring costs (i. e. the costs which are imposed for every flight), a reusable launcher requires inspection and maintenance prior to each mission and, as experience with the Space Shuttle showed, this will be more complicated and hence more expensive than for a conventional aircraft. Furthermore, it is currently expected that due to the high strains involved in launch and re-entry in combination with the need to keep the vehicle’s structure extremely light, spaceplanes will be able to make at most several hundred flights before they must be scrapped. Replacement rates and costs will therefore be higher than for airliners, where a single aircraft may undertake over 10,000 flights before having to be withdrawn from service. The operations costs for reusable systems therefore run the risk of turning out to be actually greater than for expendable rockets. The high development, infrastructure and maintenance costs mean that operating reusable launchers can only translate into attractive launch prices if they perform many flights per year. It is just like with commercial airlines, which keep their planes in the air for as many hours as possible in order to keep their costs down. A rapid turn-around is required to limit the size and hence the buy-cost of the vehicle fleet. Yet to be able to make many flights there must be a large number of customers who require many more payloads to be launched than is currently the case: today there are about 70 launches per year, although some carry multiple satellites; in contrast, on any normal day there are close to 30,000 airliners in the skies above the USA alone. But the launch market will only significantly increase in size (with space tourists and new satellite applications which are currently prohibitively expensive) if launch prices fall by a factor of 10 or so, which in turn requires efficient reusable systems with low maintenance overheads. It is a difficult Catch-22 situation: launches should become cheaper when the market is sufficiently large, but the market cannot dramatically increase until launch prices drop significantly. How the launch market will grow as a function of launch price reductions is debated heavily, and seems to be driven more by opinions than by hard statistical data.

In addition, current spaceplane designs are only capable of reaching low orbits so expendable rocket stages would still be required to boost satellites into higher orbits. And of course these ‘kick stages’ eat up payload volume and weight. Most current spaceplane concepts would not be able to place today’s telecommunication satellites into geosynchronous orbit, this being the most profitable part of the non­government satellite launch market. Spaceplanes therefore need a large new market in low orbit, something that space tourism could provide if the flights were sufficiently affordable and safe; failing that, they will have to rely upon the increased use of small satellites intended to work in low orbit. Whether new markets would be sufficient to justify the development of a reusable launch vehicle is the $10 billion – plus question that is very difficult to answer right now. Even in the mature and well understood airline market, aircraft companies are generally betting the farm when engaging in the development of large new aircraft like the Airbus 380 and the Boeing 787 Dreamliner. You can imagine what the risks will be in trying to enter a relatively new, poorly understood market like that of future space launches with projects having costs on such a scale. In addition, spaceplanes face competition from smart low-cost expendable launchers, especially at low flight rates. For instance, Reaction Engines estimates that at a flight rate of 70 missions per year, a single flight of their Skylon spaceplane would cost in the order of $30 to 40 million. Therefore in terms of cost per kilogram into orbit this means Skylon might be beaten by the SpaceX company’s Falcon Heavy expendable launcher whose development budget was much lower than that for Skylon.

An additional problem is the very long time to bring a complicated vehicle like a spaceplane into operational service: the Airbus 380 took about 13 years to develop, the Ariane 5 rocket about 12 years, and the F-22 fighter aircraft around 20 years. Any aircraft which incorporates as many new technologies as an airbreathing spaceplane will take at least two decades to advance from conceptual design to fully operational system, and that is not counting the additional time to develop and fly any sub-scale pathfinder test vehicles.

It is therefore not difficult to appreciate why spaceplane projects find it very hard to attract private investors; they are generally not interested in high-risk ventures that might deliver some unknowable return on investment after several decades. To limit the risks it seems to make sense to first build one or more smaller, less complex, less expensive demonstrators before committing to the development of a fully operational spaceplane. This philosophy appears to have been adopted in the US, where several hypersonic and scramjet test vehicles are currently being developed and flown. But while the investments are still substantial, such prototypes typically do not have any commercial use. Governments often finance at least the development and prototype phase; indeed this is how most expendable launch vehicle developments started, and is how the Concorde came into being. Government organizations could certainly play an important role in the develop­ment of the basic technology, such high-temperature materials and scramjets, just as in the early days of aviation the basic airfoil shapes were developed by NACA (the forerunner of NASA) and subsequently employed in almost all aircraft, even today. Apart from purely economic reasons the development of new strategic technology, the generation of high-quality jobs, the guaranteeing of national and especially military access to space, and indeed national prestige, can all serve as stimuli for governments to invest in high-risk technological projects such as spaceplanes. But new expendable launchers can also satisfy many of these desires at potentially lower costs and risks.

If it is difficult to close the business case for orbital spaceplanes, is it possible that a suborbital vehicle may make more sense? A smart concept that was investigated as part of the FESTIP study of the mid-1990s was that of a ‘Suborbital Hopper’ that involves a reusable vehicle which releases its payload into space at a speed just short of that required for orbit. A small rocket stage then gives the cargo the final kick to enter orbit. Such a launcher saves huge amounts of propellant by not having to boost its own weight into orbit, and is less constrained by the need to keep structure weight to the absolute minimum. In addition, such a vehicle may find profitable markets in rapid point-to-point transportation of people and time-sensitive cargo all across the world, space tourism, and undertaking short-duration microgravity and high- altitude experiments that need more time than can be provided by sounding rockets (point-to-point transportation may account for only a small fraction of the commercial aviation industry but because that industry is enormous it might still be far bigger than any short-term space launch market). In a mihtary role it could act as a strategic bomber, uncatchable spy plane, rapid-reaction satellite launcher and rapid intervention vehicle capable of delivering special forces anywhere in the world within 2 hours. This last application is currently being studied by the US Marines under the name SUSTAIN, for ‘Small Unit Space Transport And INsertion’. In fact, most work currently done in the field of hypersonic flight and propulsion is primarily for military purposes and is not intended to make space available to civilian travel and commerce. A suborbital spaceplane able to (almost) fly around the globe once would be something in between the short-range suborbital rocket planes described in the next chapter and a fully orbital spaceplane: a lower-cost, lower-risk project paving the way for a truly orbital spaceplane both in terms of technology and market development.

The ‘Astroliner’ suborbital rocket plane launch system proposed by Kelly Space & Technology in the US in the 1990s was a similar concept, with the addition of a Boeing 747 serving as a first stage. The jetliner would tow the rocket plane to an altitude of 6 km (20,000 feet) and Mach 0.8. The Astroliner would separate and shoot up to 110 km (360,000 feet) in order to release an expendable upper stage through a nose door and place several metric tons of payload into a low orbit. The rocket plane itself would continue its suborbital trajectory, re-enter the atmosphere and land on a conventional runway. The Astroliner would have jet engines for tow – flight assist and powered final descent and landing, and three Russian RD-120 liquid kerosene/liquid oxygen rocket engines for the zoom into space. During 1997 and 1998 the company conducted tests of the tow-launch concept at Edwards using a modified F-106 Delta Dart jet fighter towed behind a large C-141 Starlifter transport aircraft. Apart from this, the project does not seem to have progressed much although the concept is still advertised on the company’s website.

Most current launchers are not exactly environmentally friendly because they burn large amounts of kerosene and rubber-like solid propellants on every flight. However, since the worldwide launch rate is very low their impact when compared to airplanes or cars is fairly negligible. Several modem rockets use liquid oxygen and hydrogen as propellants for at least some of their rocket stages, the combustion of which results in nothing more than water vapor. But what if spaceplanes are launching into orbit on a regular basis? The good news is that owing to the need for high performance, these vehicles will very probably also use hydrogen as fuel and burn it with oxygen drawn from the air during airbreathing flight phases and then with liquid oxygen for rocket propulsion. They would not emit any carbon dioxide or toxic gases. However, even water vapor may not be completely harmless when emitted at massive rates: at high altitudes it may linger for a long time, and it is not yet clear what the environmental impact would be. The water condensation trails left in the sky by high-flying jets have, for instance, already been shown to have a measurable effect on the amount of sunlight which reaches the ground. Moreover, liquid hydrogen is difficult to produce; it currently requires around 15 kilowatt- hours of energy per kilogram, so the source of the energy for making the fuel becomes very important. But that is not a particular spaceflight problem, it is part of the overall clean-energy issue.

Compensating for pollution by spaceplanes might be an increase in environment­monitoring satellites as a result of a fall in launch prices, data from which may well increase our understanding of weather and climate and result in the proper measures being taken to protect our world. In addition, astronauts generally return from space deeply impressed with the notion of how small the Earth really is and how thin the atmosphere appears from orbit. Flying more people into space may greatly increase awareness of the fragility of our planet. Finally, the heavy usage of hydrogen fuel by spaceplanes may boost the world’s hydrogen industry. Spaceplanes could very well become the first large-scale commercial users of liquid hydrogen, reducing hydrogen prices and stimulating the development of efficient production, transportation and storage technologies. The economy and practicality of clean, hydrogen-powered cars could be improved by this. At the very least, spaceplane operators could incorporate energy-efficient systems and renewable energy sources into their ground operations; a new industry has the advantage that it can adopt sustainability and environmental awareness right from the start. Of course, the fact that spaceplanes would be reusable should save much energy and materials for the production of vehicles in comparison to expendable rockets, the valuable structures and other equipment of which are lost when they burn up in the atmosphere or crash into the sea.

In spite of all the potential benefits, developing and possibly flying a spaceplane or any type of reusable launch vehicle is still an (economic and technical) adventure rather than an everyday routine. And, as stated in the quote that opened this chapter, that is inhibiting success.