Space Exploration

Fire Arrows and Gunpowder Rockets

The first rocket was an arrow that did not need a bow. Chinese alchemists had written down a recipe for gunpowder by 1044, in the military manual Wujing Zongyao, and within two centuries someone had tied a bamboo tube of it to an arrow shaft and let the burning powder do the pushing. Chronicles of the Mongol siege of Kaifeng in 1232 describe fire lances and thunder-crash bombs; historians debate exactly when the fire arrow became a true rocket, propelled by its own exhaust rather than merely carrying flame, but by the fourteenth century Chinese, Arab and Indian armies were launching them in volleys from racks.

The weapon came of age in southern India. The armies of Hyder Ali and his son Tipu Sultan, rulers of Mysore, fielded rockets with iron cases rather than paper or bamboo: a tube around 200 millimetres long and 40 to 50 millimetres across, strapped to a bamboo guide stick up to a metre long, with a range of a kilometre or two. Iron held higher pressure than cardboard, so the rockets flew further and hit harder. British accounts put Tipu's rocket corps at several thousand men, and at Pollilur in 1780 a volley reportedly set off an East India Company ammunition cart and helped rout the column. After Srirangapatna fell in 1799, captured Mysorean rockets were shipped to the Royal Arsenal at Woolwich.

Did you know? William Congreve's famous rockets were reverse-engineered from the Mysorean examples sent to Woolwich; he improved the casing and the stick, but the idea of an iron rocket was Indian.

Congreve rockets entered British service in 1806. The 32-pounder version weighed about 14.5 kilograms and could reach roughly 2.7 kilometres, further than most field guns of the day, and the Royal Navy fired around 300 of them into Copenhagen during the bombardment of September 1807, when the city burned for three days. Seven years later, on the night of 13 to 14 September 1814, the rocket ship HMS Erebus fired Congreves at Fort McHenry in Baltimore harbour. Francis Scott Key watched from a truce ship and wrote of "the rockets' red glare"; the line became the American national anthem. The fort held, which tells you something: the rockets were terrifying and wildly inaccurate.

That inaccuracy is why the gunpowder rocket stalled for a century. A stick-stabilised rocket wobbles, its powder burns unevenly, and a side wind sends it anywhere. William Hale's spin-stabilised design of 1844, which vented exhaust through angled holes to make the rocket rotate like a rifle bullet, helped, but by the 1860s rifled artillery was hitting targets at longer ranges with far greater precision, and rocket troops were quietly disbanded. The deeper limit was chemical: black powder gives a specific impulse of only around 80 seconds, a quarter of what a modern liquid engine delivers, so no gunpowder rocket could ever carry much, or go high.

The gunpowder rocket proved the principle, then spent a hundred years as a signal flare and a line-thrower for lifeboats. It took a new fuel, and a few obsessive dreamers, to make it a vehicle.

Imagining the Voyage

Long before anyone could leave the planet, people were writing the trip reports. Lucian of Samosata, a Syrian-Greek satirist of the second century, sent a ship to the Moon on a whirlwind in his True History and found the King of the Moon at war with the King of the Sun over the colonisation of the Morning Star; he introduced it by promising that every word was a lie. Fourteen centuries later Johannes Kepler wrote the Somnium, a dream in which a demon carries a traveller to the Moon during an eclipse. Kepler, who had just worked out the laws of planetary motion, used it to describe how the heavens would look from the Moon, with the Earth hanging motionless in the sky and turning once a day. Published in 1634, four years after his death, it counts as the first science fiction built on real astronomy.

The seventeenth century added the vehicles. Francis Godwin's The Man in the Moone (1638) used a flock of trained geese; Cyrano de Bergerac, in a satire published in 1657, tried bottles of dew warmed by the sun and then, almost in passing, a frame hung with firecrackers that lit in stages. It was a joke, and also literature's first rocket to space.

Then came Jules Verne, who did his sums. In From the Earth to the Moon (1865) the Baltimore Gun Club casts a 274-metre cannon, the Columbiad, in the ground of central Florida and fires an aluminium projectile carrying three men towards the Moon. Verne chose Florida for its nearness to the equator; he computed an initial speed of around 11 kilometres per second, close to the true escape velocity; and in the sequel the capsule splashes down in the Pacific to be recovered by an American warship, the USS Susquehanna. A century later Apollo launched three men from Florida in an aluminium-alloy capsule at just under 11 kilometres per second and brought them home to a Pacific splashdown beside an aircraft carrier. Apollo 11's command module Columbia took its name partly from Verne's gun.

Did you know? Verne's cannon would have killed its crew in the barrel: reaching escape velocity in 274 metres means an acceleration of more than 20,000 g, when a trained human tolerates about 9.

H. G. Wells took the other route in The First Men in the Moon (1901), with a gravity-shielding substance called Cavorite and a Moon full of insect-like Selenites, and Georges Méliès turned the two books into A Trip to the Moon in 1902, fourteen minutes of painted sets ending with a capsule in the Man in the Moon's eye. Wells, not Verne, caught the next generation: Robert Goddard read The War of the Worlds as a teenager, and Hermann Oberth read Verne at eleven and spent his life checking the arithmetic.

The pattern matters because it kept repeating. Konstantin Tsiolkovsky wrote space novels alongside his equations; Wernher von Braun wrote a Mars novel in 1949; Arthur C. Clarke proposed the communications satellite in a magazine in 1945. Fiction did not just predict spaceflight, it recruited the people who would build it.

Tsiolkovsky and the Rocket Equation

Konstantin Tsiolkovsky lost most of his hearing to scarlet fever at ten, was turned away from school, and taught himself physics in Moscow's libraries, eating black bread so that he could afford books. From 1892 he taught arithmetic and geometry in Kaluga, a provincial town 150 kilometres south-west of Moscow, and in a workshop at home he worked out, on paper, how to leave the Earth. His notebook dates the key result to 10 May 1897. In 1903 the journal Nauchnoye Obozreniye published his paper Exploration of Outer Space by Means of Reaction Devices, and almost nobody read it.

What he had found is now called the rocket equation: the change in velocity a rocket can achieve equals its exhaust velocity multiplied by the natural logarithm of its starting mass divided by its final mass. Written out, Δv = ve ln(m0/m1). It follows directly from Newton's laws, and a British mathematician, William Moore, had derived a version of it in 1810 for Congreve rockets, but Tsiolkovsky was the first to point it at space and to see what it implied.

The logarithm is the cruel part. Orbital speed is about 7.8 kilometres per second, and gravity and air drag during the climb add another 1.5 or so, so a rocket needs around 9.4 kilometres per second of Δv. A kerosene engine exhausts at roughly 3 kilometres per second. Divide, take the exponential, and the starting mass must be 23 times the final mass: 96 per cent of the vehicle on the pad has to be propellant, leaving 4 per cent for tanks, engines, structure and payload. No one can build a tank that light. Tsiolkovsky's answers, in 1903, were the two that still apply. Use the fastest exhaust you can, which is why he proposed burning liquid hydrogen with liquid oxygen, a pairing not flown until 1963. And throw away the empty tanks as you go. His 1929 paper on rocket trains described stages that burn in turn and fall away, so that each later stage starts the equation afresh with a lighter vehicle.

Did you know? Tsiolkovsky built Russia's first wind tunnel in 1897, in his own house, and tested model shapes in it with a fan he made himself.

He also drew the pressurised cabin, the airlock, the spacesuit, the closed-loop greenhouse and the orbital station, and wrote science fiction, On the Moon in 1893 and Beyond the Planet Earth in 1920, to carry the ideas to readers who would never open a physics journal. In a letter of 1911 he wrote the line that every space agency has quoted since: the Earth is the cradle of humanity, but one cannot live in the cradle forever. He died in Kaluga in September 1935. Sputnik launched twenty-two years later, in the centenary year of his birth, on a staged rocket designed by engineers who had grown up on his pamphlets.

The equation has not changed. Every launch vehicle flying today, from Soyuz to Starship, is an argument with that logarithm, and staging is the only answer anyone has found.

Goddard's Liquid Fuel

On 16 March 1926, in a snowy cabbage field on his aunt Effie's farm at Auburn, Massachusetts, Robert Goddard lit a rocket fuelled by petrol and liquid oxygen. It burned for 2.5 seconds, climbed 12.5 metres, and landed 56 metres away in the frozen field, at an average speed of about 96 kilometres per hour. Four people watched: Goddard, his wife Esther with a cine camera that ran out of film, his assistant Henry Sachs and a Clark University colleague, Percy Roope. It was the first flight of a liquid-propellant rocket, and it looked like nothing: a three-metre frame of thin pipes with a small engine at the top and the tanks below. Goddard thought an engine above the tanks would make it stable, like pulling a cart; it does not, as he soon found. The date, the figures and the photograph of Goddard beside the frame come from his own meticulous diary.

Goddard, a physics professor at Clark University in Worcester, had been chasing this since a day in 1899 when, aged seventeen and sitting in a cherry tree, he imagined a device that could rise to Mars. He patented the liquid-fuelled and multi-stage rocket in 1914, and his 1919 Smithsonian paper, A Method of Reaching Extreme Altitudes, mentioned in a footnote that a rocket could carry flash powder to the Moon. The New York Times ran an editorial on 13 January 1920 jeering that he lacked "the knowledge ladled out daily in high schools", since everyone knew a rocket needed air to push against. Newton's third law says otherwise. The paper printed a correction on 17 July 1969, while Apollo 11 was on its way to the Moon.

Did you know? After a noisy 1929 flight at Auburn drew fire crews and reporters, a local headline announced that Goddard's "Moon rocket" had missed its target by 238,799 and a half miles.

The ridicule, and a ban on further launches in Massachusetts, made him secretive. Charles Lindbergh took an interest, and from 1930 the Guggenheim family funded Goddard to move to Roswell, New Mexico, where in a decade of quiet work he solved most of the problems of the modern rocket: gyroscopic steering with movable vanes in the exhaust, turbopumps, regenerative cooling of the combustion chamber, gimballed thrust. In March 1935 one of his rockets passed the speed of sound, and in 1937 another reached around 2.7 kilometres. He filed or was granted 214 patents, many of them posthumously through Esther, and they cover almost every part of a V-2 or a Saturn V.

Yet his influence on the people who built those machines was small, because he published so little. The German engineers at Peenemünde worked from Oberth's book and their own tests; when they later saw his patents, Wernher von Braun said Goddard had been ahead of them all. Goddard died in August 1945, three months after the last V-2 fell.

NASA's Goddard Space Flight Center in Maryland carries his name. The cabbage field is now a golf course with a small stone marker, which is roughly the attention he received in his lifetime.

Oberth and the Rocket Societies

The rocket entered the 1920s as a book. Hermann Oberth, a Transylvanian Saxon who had read Verne at eleven, submitted a dissertation on rockets to Heidelberg in 1922 and had it rejected; he published it himself in 1923 as Die Rakete zu den Planetenräumen, The Rocket into Planetary Space, 92 dense pages proving that a liquid-fuelled rocket could leave the Earth, carry people and survive the trip. It sold out, and it did something the earlier pioneers had not managed: it created a readership, which within a decade had organised itself into clubs on three continents, and the clubs were firing rockets.

The German Verein für Raumschiffahrt, the Society for Space Travel, was founded in Breslau in July 1927 and by 1930 had a home: the Raketenflugplatz, a disused ammunition dump in Berlin-Reinickendorf, leased by Rudolf Nebel. Klaus Riedel built the engines, Willy Ley wrote the publicity, and a nineteen-year-old engineering student named Wernher von Braun turned up to help. Their Mirak and Repulsor rockets, little more than a motor and a pair of tanks, reached a kilometre or so; by 1932 the society counted 87 flights and 270 static firings. Johannes Winkler, the VfR's first president, had flown Europe's first liquid-fuelled rocket, the HW-1, near Dessau on 14 March 1931. Then the German Army came calling, the club dissolved in 1934, and von Braun went to work for it.

Did you know? The countdown was invented by a film director: Fritz Lang added "zehn, neun, acht ..." to Frau im Mond in 1929 for suspense, and real launches adopted it.

Oberth himself had spent 1929 as scientific adviser on Lang's Frau im Mond, which showed a two-stage rocket, acceleration couches and weightlessness aboard, and he promised the studio a real rocket for the premiere. He was injured in a test and the rocket never flew, but the film reached audiences the book could not. In New York, the science fiction editor David Lasser and the journalist G. Edward Pendray founded the American Interplanetary Society on 4 April 1930; their second rocket flew from Staten Island on 14 May 1933 to about 75 metres before its oxygen tank burst. In Moscow, the Group for the Study of Reactive Motion, GIRD, formed in 1931 around Friedrich Zander, Mikhail Tikhonravov and a 24-year-old aircraft designer, Sergei Korolev. Their GIRD-09, burning gelled petrol with liquid oxygen, rose about 400 metres over Nakhabino on 17 August 1933, the first Soviet liquid rocket; that autumn GIRD and Valentin Glushko's Leningrad laboratory merged into a state institute. Britain's Interplanetary Society, founded in Liverpool in 1933 by Philip Cleator, could not fire a rocket under the Explosives Act of 1875, so it designed on paper, producing the first serious lunar lander study in 1939.

The pattern repeats across the four cities: a book, a club, a field, a crude rocket, then the state. The amateurs proved liquid rockets were real and trained the people who would industrialise them. Oberth lived to ninety-five, long enough to watch Apollo 11 launch beside his former student von Braun.

Every launch count since has begun with Fritz Lang's number ten.

The V-2

At 15:58 on 3 October 1942 an A-4 rocket rose from Test Stand VII at Peenemünde, on the Baltic island of Usedom, climbed to 84.5 kilometres and fell into the sea 190 kilometres away. Two earlier attempts that summer had failed; this one worked, and Walter Dornberger, the Army officer in charge, told his engineers that evening that they had opened the door to space travel. The machine he was describing is the one history calls the V-2, for Vergeltungswaffe, the "vengeance weapon", a name the propaganda ministry attached in 1944.

It was 14 metres tall, 1.65 metres across and weighed 12.5 tonnes fuelled, of which 8.8 tonnes were propellant: a 75 per cent ethanol-water mix and liquid oxygen, fed by a turbopump driven on steam from hydrogen peroxide. The engine produced about 25 tonnes of thrust, 245 kilonewtons, for roughly 65 seconds, by which time the missile was travelling at around 1,600 metres per second, four and a half times the speed of sound. Three gyroscopes sensed its attitude; graphite vanes in the exhaust and small rudders on the four fins steered it; an accelerometer or a radio beam cut the engine at the planned speed. The warhead carried 738 kilograms of amatol. On an operational shot the trajectory peaked near 88 kilometres and the whole flight to a target 320 kilometres away took about five minutes, arriving at supersonic speed with no warning. Every one of these figures is in the Peenemünde test records that fell into Allied hands in 1945.

Did you know? A vertical test launch on 20 June 1944 reached 176 kilometres, the first human-made object to pass the Kármán line and enter what is now defined as space.

The programme's leaders were Dornberger and Wernher von Braun, technical director at thirty. After the RAF bombed Peenemünde in August 1943, production moved underground to the Mittelwerk, tunnels in the Kohnstein hill near Nordhausen, where prisoners of the Mittelbau-Dora concentration camp assembled the missiles in cold, dust and hunger. About 6,000 V-2s were built. Historians estimate that around 20,000 prisoners died at Dora and its sub-camps, more than the roughly 9,000 people the V-2 killed at its targets, and the figures are not disputed: more people died building the V-2 than were killed by it.

The attacks began on 8 September 1944, on Paris in the morning and on Chiswick in west London at 18:43, where three people died in Staveley Road. Of around 3,170 rockets launched in anger, Antwerp received about 1,600 and London about 1,350, including a single hit on a Woolworths in New Cross that killed 168 shoppers. About a third of all launches failed. The last fell on 27 March 1945. As a weapon it was a strategic failure, costing Germany more than it cost the Allies; as a machine it was the ancestor of every ballistic missile and space launcher built in the next fifteen years, by both superpowers.

The V-2 is where spaceflight's two stories meet, the engineering and the moral ledger, and neither can be told without the other.

Paperclip and the Soviet Harvest

The V-2 did not end in 1945; it was divided. In the final weeks of the war the American, British and Soviet armies raced for the Mittelwerk, its missiles and the people who understood them, and the shares each took set the pace of the next decade. On 2 May 1945 Wernher von Braun's younger brother Magnus cycled down a Bavarian road to find an American private and surrender the team; von Braun had already chosen his captor, moving some 500 engineers and 14 tonnes of documents south ahead of the Red Army. The US Army reached the Mittelwerk tunnels near Nordhausen on 11 April, and before handing the area to the Soviet zone on 1 June its Special Mission V-2 hauled out around 100 missiles' worth of parts in 341 railway wagons.

Operation Paperclip, the programme that brought German scientists to the United States, took 118 members of the rocket team to Fort Bliss, Texas, in late 1945, von Braun among the first seven to arrive. At the White Sands Proving Ground in New Mexico they reassembled V-2s for American crews; the first launch, on 16 April 1946, failed, the second, on 10 May, reached 112 kilometres. On 24 October 1946 a V-2 carrying a 35-millimetre cine camera climbed to 105 kilometres and brought back the first photographs of Earth from space, grainy frames showing the curve of the planet. Sixty-seven V-2s flew from White Sands by 1952, carrying instruments and, twice, monkeys.

Did you know? The first launch from Cape Canaveral, Bumper 8 on 24 July 1950, was a V-2 with a WAC Corporal second stage; an earlier Bumper had reached 393 kilometres in February 1949, a record that stood for years.

Britain, with no budget for a missile programme, ran Operation Backfire instead: at Cuxhaven in October 1945, German crews under British supervision launched three V-2s into the North Sea so that the procedure could be filmed and documented. The Soviet Union, arriving at an emptied Mittelwerk, assembled what it could with the Germans who had stayed, including Helmut Gröttrup, head of the V-2's guidance group. On the night of 22 October 1946, in Operation Osoaviakhim, Soviet security troops roused around 2,200 German specialists and their families and put them on trains east. Gröttrup's group spent six years on an island in Lake Seliger, consulted but kept from the real work. Sergei Korolev's own team launched its first German-built V-2 from the new range at Kapustin Yar on 18 October 1947, flew the Soviet-built copy, the R-1, on 10 October 1948, and stretched it into the R-2 with twice the range by 1950. The Germans went home from 1951; the Soviets no longer needed them.

The accounting is lopsided but not simple. The Americans got the leaders and the paperwork; the Soviets got the factory, the production engineers and a decade of hard practice, and were first to orbit anyway. Both nations' rockets of the 1950s, Redstone and R-5 alike, were V-2s that had grown up.

Half a captured missile programme, it turned out, was enough to start a space race.

How a Rocket Works

A rocket is a machine for throwing mass out of the back as fast as possible. Newton's third law does the rest: every kilogram of hot gas flung backwards pushes the vehicle forwards with equal force, in air or in vacuum, and the vacuum is in fact better, since there is no atmosphere pressing on the nozzle exit. Thrust is the mass flow rate multiplied by the exhaust velocity. The Saturn V's first stage burned about 13 tonnes of propellant every second and threw it out at around 2.6 kilometres per second, roughly 34 meganewtons, enough to lift 2,900 tonnes at 1.2 times its weight.

The second number that matters is specific impulse, the exhaust velocity divided by the strength of gravity, measured in seconds: how many seconds one kilogram of propellant can produce one kilogram of thrust. It is set by the chemistry. Solids, as in the Shuttle's boosters, give about 240 seconds and enormous thrust; kerosene and liquid oxygen, as in the F-1, about 300 in vacuum; hypergolic fuels that ignite on contact, used by Proton and most spacecraft thrusters, a little more; methane and oxygen in SpaceX's Raptor around 350; liquid hydrogen and oxygen in the Space Shuttle's RS-25 about 452, the best any chemical engine achieves, at the cost of a fuel so light that its tanks are vast.

Getting the propellants into the chamber is the hard part. A chamber runs at 70 bar in the F-1 and over 250 bar in the Russian RD-180, so the turbopumps, turbines spun by hot gas, must push harder than that: the F-1's delivered 41 megawatts, as much as a large ship's engines. Where the hot gas comes from defines the engine cycle. A gas-generator engine, like the F-1 or the Merlin, burns a little propellant in a separate combustor and dumps the exhaust overboard, simple and slightly wasteful. A staged-combustion engine, like the RS-25 or the RD-180, sends that exhaust into the main chamber to be burned again, gaining efficiency and complexity. Raptor runs full-flow staged combustion, with every gram of propellant passing through a turbine first.

Did you know? The bell-shaped nozzle was invented in 1888 by the Swedish engineer Gustaf de Laval for steam turbines; its converging-diverging shape accelerates the exhaust past the speed of sound.

The nozzle turns pressure into speed. Gas reaches Mach 1 at the narrow throat and keeps accelerating as the bell widens, leaving at Mach 3 or more; a vacuum engine has a bigger bell because there is no air to push the plume back in. All of this fights losses: low orbit needs 7.8 kilometres per second of speed but around 9.4 of total Δv, the difference eaten by gravity during the climb, air drag and steering, which is why rockets tilt over as early as they dare.

The F-1 gave 6.77 meganewtons at sea level, the RD-180 3.83, the RS-25 1.86, a Raptor 3 about 2.75; today's heaviest rockets simply fit more of them. Each is a controlled explosion held steady for minutes, and that they work at all is the quiet miracle of the business.

Getting into Orbit

An orbit is a fall that keeps missing. Isaac Newton drew the picture in the 1680s: a cannon on a high mountain fires faster and faster, each ball landing further away, until one goes so fast that the ground curves away as quickly as it drops, and it comes round to hit the gunner from behind. Nothing is holding a satellite up. The International Space Station, 400 kilometres above the Earth, is falling the whole time; it simply moves sideways at 7.7 kilometres per second, and the planet curves out of its way. The weightlessness aboard is the weightlessness of a lift whose cable has snapped.

Height is cheap and speed is dear. Reaching 100 kilometres, the Kármán line that the records body uses as the edge of space, takes about 1.4 kilometres per second straight up; orbit takes 7.8 sideways, and that speed is 95 per cent of the energy. That is why a rocket climbs only briefly before tilting over in a gravity turn, trading altitude for velocity as soon as the thick air is behind it.

Did you know? The Earth's spin gives an eastward launch from the equator a free 465 metres per second, so Europe's spaceport at Kourou, 5 degrees north, can lift a tonne or two more to geostationary orbit than Baikonur at 46 degrees.

That rotational bonus explains the map. Kourou in French Guiana, Cape Canaveral at 28 degrees, Sriharikota at 13, Wenchang at 19: launch sites hug the equator and fire east over water. Russia's Baikonur, at 46 degrees in Kazakhstan, gets only 320 metres per second from the spin, and Plesetsk, at 63 degrees, about 210, and the ISS orbits at 51.6 degrees because that is the lowest inclination Baikonur can reach without dropping spent stages on China.

The orbit chosen depends on the job. Low Earth orbit, below 2,000 kilometres, is for stations and most satellites; sun-synchronous orbit, a near-polar track at around 700 kilometres passing each place at the same local time, is for Earth observation; geostationary orbit, 35,786 kilometres up over the equator, is where a satellite's 24-hour period matches the Earth's spin and it hangs over one spot. The Soviet Molniya orbit, a 12-hour ellipse tilted at 63.4 degrees, lingers over the Arctic, which geostationary satellites cannot see. Reaching it is a two-step: the launcher puts the satellite into a long transfer ellipse and at its far end the satellite's own motor rounds it off, the Hohmann transfer Walter Hohmann worked out in 1925, costing about 2.4 kilometres per second more.

What goes up comes down, slowly. Even at 400 kilometres there is enough air to drag the ISS down by about 2 kilometres a month, so it is reboosted a few times a year; Sputnik lasted 92 days, Skylab fell on Western Australia in 1979 and Mir was steered into the Pacific in 2001. Above about 1,000 kilometres a satellite will stay for centuries, which is the beginning of the space-junk problem. Orbit is not a place but a speed, and every launch is an argument about how to buy it cheaply.

Korolev and von Braun

The space race was run by two engineers who never met and could not have been introduced. One was a name the whole world knew: Wernher von Braun, baron, television presenter, father of the Saturn V. The other was a state secret. Until the day after he died in January 1966, Soviet newspapers called Sergei Korolev only the Chief Designer, and when the Nobel committee asked whom to honour for Sputnik, Nikita Khrushchev replied that the achievement belonged to the whole Soviet people.

Korolev, born in Zhytomyr in 1907, was an aircraft designer who led GIRD's first liquid-rocket flight in 1933. In June 1938, during Stalin's purges, he was arrested on fabricated charges, beaten until his jaw broke, and sentenced to ten years. He spent the winter of 1939 to 1940 in the Maldyak gold mine in Kolyma, where scurvy took most of his teeth and nearly his life; his mother's letters and the aircraft designer Andrei Tupolev got him moved to a prison design bureau, a sharashka, where he worked under guard until 1944. Sent to Germany in 1945 to study the V-2, he built from the R-1 copy the R-5, the first Soviet nuclear missile, and then the R-7, the 280-tonne, 32-engine rocket that flew on 21 August 1957, launched Sputnik six weeks later and Gagarin in 1961, and whose descendants still carry crews to the station. He ran the programme on charm, intimidation and a heart weakened by the camps, and died on an operating table at 59.

Did you know? The R-7's shape, a core with four strap-on boosters, exists because Korolev was told to carry a three-tonne hydrogen bomb before the physicists shrank it; the overkill became the space rocket.

Von Braun, born in Wirsitz in 1912, joined the Nazi Party in 1937 and the SS in 1940, rising to major; the records show he visited the Mittelwerk and knew the prisoners' conditions. He said afterwards that he could have done nothing; historians judge that he chose not to try. A brief Gestapo arrest in 1944, for talking about spaceflight rather than weapons, and his surrender to the Americans let him be recast. In the United States he built the Redstone, the Jupiter and the Juno I that orbited Explorer 1 on 31 January 1958, after the Navy's Vanguard had exploded on television. He sold space to the public through Collier's and three Disney television films from 1955, the first of which, Man in Space, drew around 40 million viewers. From Huntsville, Alabama, he led the Saturn V and watched it fly thirteen times without failure.

Each knew of the other and saw the race as a duel: Korolev's engineers studied the Saturn plans in the open American literature; von Braun lobbied Washington after Sputnik by pointing at a rival he could not name. Korolev's death left the Soviet Moon rocket in weaker hands; von Braun left NASA in 1972 when the Saturn line was cancelled, and died in 1977.

The race was between two men as much as two systems, and the one who won it is still the harder to judge.