FROM THE MAGAZINE
April 2015 Issue

Everything You Need to Know About Flying Virgin Galactic

More than 700 people have paid up to $250,000 for a ride on Richard Branson’s Virgin Galactic. Can he guarantee a safe landing?
This image may contain Richard Branson Airport Human Person Vehicle Transportation Airplane Aircraft and Airfield
Richard Branson in Mojave, California, in 2010. Behind him, SpaceShipTwo hangs from the twin-fuselage mother ship, WhiteKnightTwo.Photograph by Jonas Fredwall Karlsson.

I. The Unthinkable

Commercial passenger service to space is a difficult proposition. To succeed, it has to contend with the pull of gravity, violent rocket-propelled accelerations, heavy vibrations, supersonic speeds and shock waves, vertical climbs, the lethality of the outside environment, and the problems of deceleration and heating during re-entry into the atmosphere. It has to do this safely, reliably, repeatedly, and perhaps profitably, while carrying ordinary passengers in ordinary clothes, who, if they are traveling point-to-point, will want to bring along ordinary luggage as well.

We are far from that now. But by last fall, Virgin Galactic, a self-described “spaceline” founded by the flamboyant British entrepreneur Richard Branson and based in Mojave, California, was nearing the moment when it could offer rides to as many as six passengers at a time. In terms of space travel, these rides were to be baby steps—at best, a brief, straight-up excursion above the so-called Kármán line (100 kilometers high, in near-vacuum conditions, where space is somewhat arbitrarily said to begin), followed by a straight-down re-entry into Earth’s atmosphere and a gliding return to the airport of origin. There had been technical setbacks, impossible promises, and all too much hype. Nonetheless, more than 700 people had bought tickets, most recently at $250,000 apiece, and after a decade of development and $500 million of investment, the machines were impressive and real.

There were two of them, constructed of carbon composites, and each for the moment was one-of-a-kind. The first, known as WhiteKnightTwo, was a strange-looking, twin-fuselage mother ship with a 140-foot wingspan and four turbofan engines—a heavy lifter designed to suspend a rocket ship between its fuselages, take off with it from a runway, and fly it into the thin air at 47,000 feet for a horizontal launch. The rocket ship itself, known as SpaceShipTwo, was a twin-tailed, stubby-winged aircraft the size of a business jet, built around a hybrid rocket engine containing enough fuel for a one-minute burn—sufficient to thrust the aircraft to a speed of Mach 3.5 (about 2,500 miles per hour) in a vertical climb and project it into space, from which it would return without power, as a glider. It had a two-pilot cockpit and a cabin large enough for six reclinable passenger seats, as yet to be installed.

At the core of the design was an unusual twin-boom tail that straddled the rocket motor. The tail had vertical stabilizers and various movable control surfaces, and operated more or less conventionally in regular flight. What made the tail unusual—highly unusual—was that the entire twin-boom apparatus was hinged. In regular flight it was held in position by mechanical locks, but it could be unlocked and then raised to a 60-degree angle overhead—nearly perpendicular to the fuselage. This raising of the tail was meant to be done in the thin air of near space, where aerodynamic loads are small. The purpose was to control the subsequent atmospheric re-entry by forcing the fuselage and wings into a flat attitude, like a pancake or belly-flop in relation to the fall. The resulting drag would tame the descent speeds and distribute the re-entry heat across large surface areas, rather than concentrating it on narrow leading edges. In short, raising the tail would keep the ship safe during re-entry. New terms for the invention had been coined: once raised into the upright configuration, the tail was said to be “feathered,” and descents from space were said to be “feathered re-entries.” Preliminary low-altitude test-flying had shown it to work well. The spaceship had not yet flown into space, but there was good reason to believe that, once it did, it would safely negotiate the dangers of coming back down.

Read more: Subscribe now for access. The full issue is available March 11 in the digital editions and March 17 on national newsstands.

Confidence was all the greater because these aircraft—the mother ship and the spaceship—were not entirely new inventions. They were derivatives of the aircraft that in 2004 had won the $10 million Ansari XPrize, awarded to the first privately funded venture to put a human being into space. Those earlier aircraft had been built and flown by Scaled Composites, a Mojave-based company founded by an aeronautical engineer named Burt Rutan, who is famous for his innovations. As for Branson, he was a space enthusiast. He had registered the name Virgin Galactic more than a decade earlier, had been looking for the right technology, and was impressed by the tail-feathering idea. As a result, his Virgin Group helped to fund the final XPrize flights. The winning profile involved an air launch at altitude, a horizontal acceleration followed by a pull into a still-accelerating vertical climb, and an equally vertical descent under feather. Thrilled by the success, Branson decided to offer the same ride to the general public, and he turned to Rutan to scale up the XPrize designs.

SpaceShipTwo being carried aloft.

By Mark Greenberg/Virgin Galactic/Getty Images.

Branson’s career has been checkered and controversial. He has had some notable successes (Virgin Records, Virgin Atlantic) and some very public failures (Virgin Cola, Virgin Racing, Virgin Brides). Critics note that his efforts as a private entrepreneur have often involved backroom leveraging of public subsidies and infrastructure. He has embraced many causes, not always for the long haul. But judging from the investment of time and money, his commitment to Virgin Galactic is not a passing enthusiasm, even though as a purely business venture it seems to make no financial sense. Branson recently described his thinking to me over lunch in Toronto, where he had gone to advocate for the decriminalization of drugs. He said, “Like with any venture I’ve done with Virgin, I’ve never gone into it thinking, How can I make lots of money? I’ve gone into it because I want to create something. And then I try to make it pay its way. You never know where these things are going to take you. Quite often you head down one path with a bit of a dream, and then you find there are about seven other paths that open up.” The idea, at the outset, was to sell the flights as thrills: rocket rides capped by ultra-high views and the chance to float around weightlessly for several minutes, followed by the experience of a high-G re-entry, a rollicking Branson-style hangar party once back on the ground, and bragging rights ever after. As Branson put it, “The thinking was to create a spaceship company that could give people the most memorable day in their lives.” To publicize the effort, Branson had himself photographed in a NASA-style space suit, though no such space suit would be needed on the flights he proposed.

Branson expected that scaling up the XPrize designs would be a relatively simple matter. It was not, because scaling up turns out to be difficult, and also because these were to be flights offered not to fearless test pilots but to a well-heeled public that was bound to worry, as Branson put it, “about the one-way-ticket thing,” by which he meant dying. Designing a new, high-powered rocket motor proved to be especially difficult, presenting problems with reliability, predictability, vibrations, and thrust. In 2007, during a cold-flow test—a test that does not involve actual ignition—an exploding oxidizer tank killed three Scaled Composites employees and seriously injured three others. Branson told me that Rutan was devastated by the deaths and afterward personally retreated from the project. Meanwhile, as the spaceship evolved, it grew heavier than intended (as aircraft in design typically do), raising questions about future payload and altitude capabilities. Nonetheless, by last fall the engineers at Scaled Composites, which is now a subsidiary of Northrop Grumman, believed that SpaceShipTwo was ready for the fourth in a planned series of seven rocket-powered test flights.

On the morning of October 31, the mother ship, with its attached spaceship, took off from the Mojave Air & Space Port and with special permission climbed to the southeast, into the vast restricted airspace around Edwards Air Force Base. It was a typical desert day. The sky was partially covered by broken thin clouds, clearing over Mojave. The visibility was unlimited. In command of the heavy lifter was an affable Scot, David Mackay, 58, a former R.A.F. test pilot and Virgin Atlantic captain who lives for stick-and-rudder flying. Mackay had flown the spaceship’s previous test flight, nine months earlier, and after a 20-second burn of the rocket motor he had achieved Mach 1.4 and 71,000 feet, about twice as high as airliners fly. The goal of the current flight was to test a new fuel that promised smoother performance, and to push the aircraft faster and higher than it had gone before, though not yet into space. On this flight, as on the previous two, the tail-feathering configuration was to be used during the descent.

In the cockpit of the spaceship itself were two Scaled Composites test pilots, both of them engineers: Michael Alsbury, 39, who was serving as co-pilot, and Peter Siebold, 43, who was in command. They wore parachutes, but as is common in civilian prototype test-flying, they had no ejection seats. The cockpit was quiet except for the muted rush of the atmosphere outside. Far below them the desert stretched to the horizons. After a laborious climb, Mackay delivered them to a pre-determined waypoint at 46,500 feet. In the spaceship the pilots went through checklists using the standard call-and-response: “Speed brake.” “Off.” “Roll boost.” “On.” “Isolation valve.” “Closed.” Siebold was on the controls. At approximately 10:07 A.M., Alsbury armed the release. A crew member in the mother ship dropped them clear.

Three seconds into the drop, with the nose approximately level, Alsbury toggled two switches that lit the rocket. The effect was dramatic, a powerful shove that started as a kick and did not let up. The spaceship accelerated fast horizontally, pushing the pilots against their seatbacks with a force of three Gs. Nine seconds later, still accelerating at the same aggressive rate, the vehicle passed through Mach 1, the speed of sound—about 660 miles per hour at that altitude. Then the unthinkable happened. Sixteen seconds after the drop, 13 seconds after rocket ignition, and 4 seconds after exceeding the speed of sound, SpaceShipTwo suddenly came apart, exploding into pieces that powdered the sky. Michael Alsbury was killed instantly, leaving behind a wife and two children. Peter Siebold somehow parachuted to safety.

Richard Branson got word of the accident by e-mail in the British Virgin Islands, where he resides, and hired a jet to rush him to Mojave. He told me that when he landed he found that the hundreds of people on the intertwined teams at Virgin Galactic and Scaled Composites were milling around in grief and shock. Branson vowed publicly that the project would go on. A second spaceship was already being built. For now, the larger context had to be kept in mind. Unfortunate as it was, this had been an accident during the always risky process of developmental flight-testing, with willing and informed test pilots aboard; furthermore, as a test bed the aircraft had been unusually well instrumented and had sent a wealth of data back to the ground before the telemetry stopped. Investigators from the National Transportation Safety Board had arrived in Mojave and were going to have a relatively easy time; though the final report would take perhaps a year to complete, the basics were likely to be known within days. Virgin Galactic made the standard futile plea to avoid speculation in the meantime.

Speculation in the press was rampant nonetheless, and some of it was uninformed. The most prominent theory was that the rocket engine had exploded, though it had performed well during tests on the ground. An explosion was certainly within the realm of possibility, but in Mojave it was quickly known that the engine had hit the ground intact, as had the fuel tanks. The details seemed not to matter. Time magazine published an editorial titled “Enough with Amateur-Hour Space Flight,” suggesting a connection between the in-flight breakup and what the author claimed to be Branson’s arrogance and technical ignorance, and deploring the infiltration of dilettantes into the space business. The critique ignored the expertise that resides within Scaled Composites and Virgin Galactic. Amateurs? A number of team members had come directly from high positions at NASA after the shuttle program ended, and plenty of others had extensive aerospace experience as engineers, flight-operations managers, and test pilots. As for Branson, he has been a hands-off financier, helping absorb the huge costs and delays and keeping the project alive. But Branson is an anti-Establishment figure, who in all his endeavors—Virgin Mobile, Virgin Atlantic, and the rest—habitually pushes in front of the cameras, proclaiming that he can do better than others for less. Apparently his posturing and self-promotion can grate on people sometimes.

In any case, what seems to have taken the spaceship down was not ego but pilot error. The telemetry soon showed it conclusively. The error was related to the feathering mechanism—the all-important, drag-inducing re-entry device. Scaled Composites’ engineers had decided on a two-step procedure for deploying it. First, the pilots would pull a handle to unlock the hinged tail early during the rocket burn, when the ship was at Mach 1.4, and at a low enough altitude to abort the ascent into space in the unlikely event that the unlocking failed. At Mach 1.4, the aerodynamic pressures would be downward on the tail, keeping it conventionally aligned, flush with the wings and the fuselage. That was the calculation, and it included ample safety margins. Second, after the rocket burn was complete and the aircraft in its vertical climb had slowed to aerodynamic speeds of 30 miles an hour or less, the pilots would pull another handle to raise the feather—a 20-second endeavor in thin air that would raise the tail nearly to perpendicular. The two-step procedure had worked as predicted on previous test flights.

Branson, aviation designer Burt Rutan, and Microsoft’s Paul Allen watch an XPrize-winning flight, 2004.

By Robert Galabraith/Getty Images.

This time, however, was different. Eight seconds after igniting the rocket, while still in horizontal flight, and well short of the Mach 1.4 threshold, the co-pilot, Michael Alsbury, reached forward and prematurely unlocked the tail. This was seen on a cockpit video transmitted to the ground. Why he did it may never be known. Unlocking the tail was a checklist item, requiring call-and-response. Alsbury had rehearsed it with Siebold in a simulator dozens of times, and as a pilot he had a reputation for flying by the book. Unlocking prematurely was certainly an exception. The problem was that the rocket ship just then was approaching the speed of sound. Complex shock waves were developing across the airfoils, and aerodynamic pressures were for the moment upward rather than downward on the tail. Siebold was too busy flying to notice what Alsbury had done. There have been reports that Alsbury may have understood his error and fumbled to shut off the propulsion, but this seems unlikely because the normal first reaction would instead have been to relock the tail. Whatever the truth, it was too late. A few seconds after Alsbury pulled the unlock handle, the tail reared up. The spaceship pitched up violently in response and was broken apart by positive G-loads beyond its limits. The fuselage then pitched down and tumbled head over heels, experiencing negative G-loads of sufficient force to rip Siebold’s seat from the floorboards and spit him through a gap that had appeared overhead as the top of the cockpit tore away.

Alsbury’s body would be found still strapped to his seat amid the main wreckage, which lay in a five-mile swath across the desert floor. Siebold was thrown into the clear cold blue and lost consciousness. He fell for at least 25,000 feet, then came to and seemingly had the presence of mind to unbuckle from his seat, allowing his parachute to deploy, which it soon did automatically. A chase airplane circled him as he floated down, and he waved to indicate that he was alive. He suffered injuries including a broken shoulder, and after a hospital stint during which he was checked for the possible consequences of high-altitude exposure, he returned to work to push the project forward.

II. Chasing the XPrize

Burt Rutan, the ferociously inventive engineer behind Scaled Composites, had retired from the company three years before the accident and was at his new house, overlooking Coeur d’Alene Lake, in Idaho, when he heard belatedly that the breakup had occurred. In private, he was angry that he had not been told sooner. He was also angry that Alsbury, one of his protégés, had made such an avoidable mistake. Rutan was not insensitive to the tragedy—he can be a very sensitive man—but the accident seems to have evoked in him a feeling of betrayal. While he was at the company he had relegated the Virgin Galactic work to his deputies and had not played an active role in the engineering of the aircraft or the spaceship, but it was obvious that they were the children of his XPrize designs and therefore reflected on his legacy.

Rutan was raised in the 1950s in a small town called Dinuba, in California’s San Joaquin Valley. His father, a dentist, was a self-reliant man and the proud descendant of wagon-train pioneers. As a child, Rutan began to design and build balsa-wood model airplanes. He learned to fly, and soloed in a two-seat Aeronca Champ when he was 16, in 1959. His older brother, Dick, who would later set distance records by piloting some of Rutan’s designs, went off to fly fighters for the U.S. Air Force in Vietnam. Rutan took a more cerebral approach and studied aeronautical engineering at Cal Poly, in San Luis Obispo, from which he graduated in 1965. He hired on at Edwards Air Force Base, in the Mojave Desert, and spent the next seven years as a civilian flight-test engineer, strapped into airplanes behind test pilots to gather knowledge about new and existing designs. He helped to tame certain stall characteristics of the F-4 Phantom, and for this he earned an air-force medal.

But, for Rutan, that was not enough. While still in college, he had begun to design a strange-looking two-seat airplane with a pusher prop and a horizontal stabilizer on the nose—a configuration meant to be stall-proof and stable, and therefore especially safe. It was an ugly little thing, made out of wood and fiberglass, and very slow, but it was highly inventive aerodynamically. Rutan called it the VariViggen—the reference being to the Saab 37 Viggen fighter, which has some design similarities. In 1972 he flew the VariViggen for the first time, and it handled well. As a homebuilt airplane it was not vouched for by the F.A.A. and was officially classified as experimental. This is a standard category for homebuilt aircraft—even those constructed, as most are, from store-bought plans and kits. Such airplanes cannot be mass-produced or flown commercially, but they can be built (and later bought and sold) free of the governmental oversight that characterizes conventional aircraft development. Partly as a result, they tend to perform better than manufactured airplanes do, and at lower cost. The advantages generally come at increased risk, but the market for homemade airplanes is large—there are perhaps 150,000 potential do-it-yourselfers in the United States alone. It was these people whom Rutan went after. In 1974, having left his job at Edwards, he set up shop in a shed at a civilian airport in nearby Mojave and began selling plans for the VariViggen.

The town of Mojave is hardly more than the slowing of a desert highway. With its liquor stores and flimsy motels it looks like the setting for a film-noir crime spree. It is dwarfed by the adjoining airport, a former Marine Corps flight-training field with long runways, big hangars, and square miles of empty ramp space. The airport lies just outside the restricted airspace around Edwards Air Force Base—a huge piece of sky called R-2508, where high-performance flying can be done. Recognizing these special attributes, in the early 1970s two private pilots with libertarian leanings—a Kern County farmer turned airport manager named Dan Sabovich and Congressman Barry Goldwater Jr.—arranged to declare the airport a civilian test-flight facility, an aviation oasis where regulators would not interfere. Sabovich intended to facilitate experimentation by providing shop space at low cost and then staying out of the way—and that is what he did, giving birth to what has since grown into the greatest hotbed of innovation that aviation has ever known. Rutan was among the first tenants, at 20 cents per square foot. He called his enterprise the Rutan Aircraft Factory.

A mock-up of the spaceship's tail raised in the all-important “feathered” position. By Jules Annan/Uppa/ZumaPress.com.

Over the next few years he designed and flew a series of unconventional airplanes, primarily two-seaters made of composite materials. All of them were test platforms, one-off prototypes, some of which, once refined, formed the basis of plans that Rutan offered to the homebuilt-aircraft market. The business was a success—the plans sold by the thousands. The downside was that hundreds of the finished products soon flew, and though they were designed to be unusually safe, a normal number of them crashed, sometimes with fatal results. The accidents forced Rutan to confront two related truths in aviation: that even the best-designed airplanes are never foolproof, and that claims to exceptional safety are never wise. These are truths that Virgin Galactic might bear in mind. For Rutan, the libertarian oasis in Mojave turned out to be a mirage. Lawsuits forced him repeatedly into court to defend his designs. In 1985 he shut the whole thing down, scrapped the Rutan Aircraft Factory, and never again sold plans to the public. He had already founded a new company, Scaled Composites, to design and build mission-specific prototypes and engineering models for the aerospace industry, and this is what he did for the next 26 years.

Only one of those designs ever went into production—the Beechcraft Starship—and it was soon pulled from the market. All other Rutan designs were experimental, including many made for Department of Defense contractors. The truth when it came to libertarianism was that big government had arrived, and with it the requirement to safeguard state secrets, the ultimate form of regulation with which Rutan complied. But he prospered, as did the airport, and Scaled Composites grew to several hundred employees. In some cases the quest for invention bordered on the meaningless. There was, for instance, the desire of his brother Dick, who had retired from the air force, to be the first to fly around the world without refueling. It was an empty stunt, but Rutan built the airplane to let it happen—essentially a large powered sailplane with plenty of fuel—and in 1986 the flight of the Voyager was made from Edwards to Edwards while crossing the equator twice. The flight took nine days and proved to be difficult and risky, but the publicity was huge, with thousands in attendance for the landing, and extensive television coverage. Afterward, Dick went off on the dare-to-dream circuit, and Rutan became widely known, legitimately, as the most original airplane-maker of recent times.

In the 1990s, space beckoned, and Rutan began to think about sub-orbital space travel, only to run up against the very reasons it is not done—safety, expense, and the lack of appropriate materials and technologies. Around 2001 he set aside the point-to-point ambition—getting space travelers from Los Angeles to Tokyo, for instance—and settled on the more modest idea of going after the XPrize. The challenge was very specific: to build a vehicle capable of carrying people, however briefly, above the 100-kilometer Kármán line, and to send it there twice within two weeks. Rutan decided on an approach similar to that used by the X-15 rocket planes of the 1960s: an airborne launch, a rocket-propelled climb, and a gliding return to the runway. The craft he designed would be hand-flown into space and back. It would weigh about 8,000 pounds fully loaded, one-quarter the weight of the X-15 but with a thrust-to-weight ratio of two to one, which was about the same. Unlike any other spaceflight that has ever been attempted, it would not describe an arc but, after the horizontal launch, would go straight up, hit zero velocity at the apogee, and fall straight back down. Atmospheric re-entry was the big worry. Rutan had been among the search party at Edwards in 1967 when an X-15 disintegrated at Mach 4, killing the pilot because he had gotten the re-entry angle wrong. Rutan was determined not just to slow the re-entry speeds but to take the angle out of human hands, so that in this critical phase the aircraft would sort itself out aerodynamically, and no piloting skill would be required. A tail-feathering configuration promised to be the solution. And it was.

Microsoft co-founder Paul Allen, who is an aviation enthusiast, was convinced by the plan. He flew to Mojave in his private Boeing 757 and offered to fund the effort—ultimately for $28 million—in return for a majority stake in the intellectual property and the licensing fees that might result. He and Rutan formed a company to own the project, and gave the contract to build and fly the aircraft to Scaled Composites. The close interweaving of nominally distinct companies has long been a feature of Rutan’s operations and has affected Virgin Galactic as well. In any case, Allen was less interested in the two formal XPrize flights than in the full-altitude test flight that would precede them: he wanted to be the man behind the first privately funded human spaceflight in history. That flight, after 14 previous test flights at lower altitudes, occurred on the morning of June 21, 2004, in front of thousands who had gathered at the Mojave airport to witness the event.

The test pilot was Michael Melvill, an unassuming 63-year-old immigrant, originally from South Africa, who 30 years earlier had completed the first plan-built VariViggen, and then had gone to work for Rutan to help run the shop. Melvill was unusually good with his hands, and useful therefore in the building of prototypes, but his greatest talent turned out to be in test-flying airplanes. Indeed, he was a genius at it—as good as the best have ever been. As a former flight-test engineer, Rutan recognized this. Melvill was an ordinary-seeming man in person, and he did not fly simulators very well (that was for younger pilots and video-gamers), but if you put him into an experimental airplane in flight he became transformed, able to execute flight-test plans with precision, to remain unflappable no matter what the circumstances, and to come home time after time with the data required. It helped that he was also intensely devoted to Rutan and willing to risk his life on the master’s behalf.

He certainly did that now. The renowned aviation writer Peter Garrison, himself an accomplished aerodynamicist and aircraft builder, described the first spaceship as hardly more than a tiny hot rod consisting of an airtight cabin glued to the front end of a homemade rocket. With eyes to the future, Rutan called it Spaceship One. Melvill dropped with it from the mother ship at 47,000 feet and lit the fire. He had done this once before, on an earlier test flight, and had been shocked by the forces, the vibrations, and the howling of the rocket motor; this second time, he told me, the experience was better because he knew what to expect. But, suddenly, as he was pulling through 60,000 feet, wind shear rocked the spaceship 30 degrees wing-down to the right. Surprised, Melvill countered with excessive controls, and the ship responded by rolling 90 degrees wing-down to the left, and then, as he corrected again, the same to right. Melvill finally got the wings level, and under heavy vibrations and G-loads continued into a vertical climb. He shut off the engine 76 seconds into the burn while climbing through 180,000 feet at Mach 2.9, and kept climbing purely as a projectile. Working now in weightless conditions, he raised the tail on schedule into the feathered position. The ship continued to shoot upward through the vacuum outside.

SpaceShipTwo pilot Mike Alsbury (second from left), who would be killed in the accident, and David Mackay (far right) before a test flight, 2013.

By Jason DiVenere/Virgin Galactic/Getty Images.

Then came some bad news: when Melvill tried to trim the tail’s twin horizontal stabilizers to adjust the aerodynamic surfaces for the coming glide, he discovered that the trimming mechanism had failed, leaving the tail configured in a way that would cause the spaceship to spin out of control catastrophically during the descent. Watched over by mission controllers on the ground, who were observing the telemetry, Melvill tried to remedy the problem using a backup system, but to no avail. It was a death sentence. He later told me, “As you can imagine, it made me feel pretty bad.”

That was his mood when he crossed the Kármán line and entered space for the first time. With nothing better to do, he took some M&Ms from his pocket and set them free to float around. He hit the apogee at 328,491 feet and then started down toward self-destruction. As the aircraft descended into the upper atmosphere and the G-load began to grow, he reached up to the backup trim switch and nudged it again, though without hope. To his surprise, this time it functioned correctly and eased the angle of the errant stabilizer. He saw the indication on an instrument. The mission controllers saw the indication, too, and advised him to use the switch again—not to the full extent necessary to normalize the trim, because of concern about inducing another failure, but sufficiently to provide perhaps for a controllable glide. This is what he did, fighting powerful tendencies to roll to the right and countering in the end with nearly full-left stick, but succeeding with a smooth landing back at Mojave.

At the airport the crowd cheered. Paul Allen was satisfied with his investment. Melvill had experienced 3.5 minutes of weightlessness, had spent 10 seconds in official space, and had endured a feathered re-entry at five Gs without burning up. The design had worked, and Melvill had become the first commercial astronaut.

Rutan gave him a sign to hold up for the cameras. It was provided by an Arizona radio host and activist from the Western Libertarian Alliance, Rutan’s fellow travelers, who were well represented in the crowd. The sign read, SPACESHIPONE, GOVERNMENTZERO. The message ignored the fact, as Rutan has done on other occasions in public mockery of NASA and big government, that Scaled Composites had been living on Pentagon dollars for years. The euphoria was overwhelming. There was even talk of private flights to a new space hotel, where guests would have bubble-glassed rooms for performing weightless sex while orbiting the earth. But Melvill’s flight had been a very near thing.

The same was true of the two successful XPrize flights several months later, in the autumn of 2004. The first was flown by Melvill again. As a result of design errors related to wing configuration, the spaceship became unstable while accelerating through Mach 2.7 in the vertical climb and began to roll rapidly to the right. Melville was able to slow the rolls by using conventional aerodynamic controls as he spun upward into the near vacuum of space. After he extinguished the rocket and feathered the tail, he managed to stop the roll entirely by employing the reaction-control system—little nozzles that regulated the ship’s attitude in space by shooting out streams of compressed air. In the process he consumed the entire compressed-air supply, leaving the ship steady but uncontrollable in space. He rose through the Kármán line in that condition, and reached the apogee at 337,700 feet, about 9,000 feet higher than he had gone before. The feather sorted things out on the way down, though it swung him through pendulum oscillations as it did. After he landed and could speak to Rutan in private, he expressed disappointment at having expended all the compressed air. He told me that Rutan answered, “No, don’t you see? Before I was just guessing how much we might need, and now it turns out I was right!” Yes, but with no safety margin built in. Even Melvill was taken aback.

The second and final XPrize flight was flown by another test pilot, Brian Binnie, and though it was equally violent, it was less eventful because of modifications to the climb profile. Binnie reached 367,500 feet, 13,000 feet higher than the X-15 record, and returned to land safely in front of the usual ecstatic crowd. But Rutan knew how close to the edge even that flight had skated. Having escaped space with his reputation intact, he vetoed the idea of yet another flight, and Melvill readily agreed. He told me later he was just glad to have finished the program alive.

III. Into the Sky

Richard Branson went to Mojave to witness the final XPrize flight. He had joined the effort and begun providing operational subsidies in return for having a Virgin Galactic logo displayed on the tail, and was so confident of its success that he had already signed a contract with Paul Allen to adopt the technology under license for his space-touring venture. Allen was there, too. As the spaceship rocketed upward, Branson turned to him and said, “Paul, isn’t this better than the best sex you ever had?” Allen did not answer “Yes” or, better yet, “No.” What was going through his mind, he later wrote, was “If I was this anxious during any kind of interpersonal activity, I couldn’t enjoy it very much.” Branson and Allen are certainly different men. Allen says he will not go into space, because he has learned to be afraid. Branson vows not only to be on the first commercial flight but to bring his children as well.

Despite the in-flight breakup last October, there is reason to believe that Branson will have his chance, and that others will follow along. Only about 20 passengers have canceled their tickets. The development work in Mojave continues unabated. After the second Virgin spaceship (now being built by Virgin Galactic itself) is completed, flight-testing will resume. Commercial flights could begin toward the end of 2016. They will take off and land from a “spaceport” in the desert near Truth or Consequences, New Mexico, which the state government is funding at a cost of $209 million. The passengers will participate on the legal basis of informed consent, much as people do in homebuilt aircraft, without the assurances that apply to airline travel. But informing passengers is not quite the same as warning them, and warnings don’t necessarily sink in. Branson is dealing with the ordinary lay public here but offering an experience that is different from the usual product. Some passengers on the list have already asked if they can bring music along, or cameras so they can take pictures of themselves, and others have asked about WiFi so they can tweet during the flights. Such people are woefully unprepared for this rocket-ship ride.

Branson’s spaceship is not an airliner. It is smoother and more stable than Rutan’s earlier XPrize machine, and less susceptible to upsets, but it is a radical design nonetheless, and it will be flying a profile that is equally extreme. Virgin Galactic pretends that the flights can become routine, but given that all four of the pilots currently slated to make the first passenger runs have high-performance military flight-testing backgrounds, the company seems to have recognized the challenges it will face in keeping the customers alive. I asked David Mackay how the spaceship handles. In the understated manner of test pilots, he said, “For a vehicle that launches at 120 knots, accelerates to Mach 3.5, goes up to the altitudes it goes to, then folds itself to re-enter, it flies surprisingly well.”

The spaceship disintegrates, 2014.

From Reuters/Corbis.

It is possible to fill in some details of future flights when passengers will be aboard. The cockpit is open to the cabin. The pilots sit side by side at the controls. The passengers—as many as six—sit in two single-seat rows behind them, beside large circular windows, with other windows overhead. Their seats are upright for the launch and climb but can be rotated backward into a reclined position to counter the effects of G-loads during the re-entry. It will be helpful if the passengers have gone through some high-G and zero-G training in advance, in addition to the two or three days of preparation that Virgin Galactic expects to provide in New Mexico during the run-up to the flight. Over many years as an aerobatic flight instructor, I observed that people quickly learn to handle loads of up to three Gs, but that, depending on their physical condition and seat position, some get tunnel vision and begin to black out above that. Zero and negative Gs present no such risk, but, like flying upside down, take longer to get used to—usually about five hours of aerobatic training in flight until the vomiting stops.

So, here’s the scene, and it starts off clean. Each flight should last about an hour and a half in total. The passengers sit strapped in and upright wearing padded helmets and flight suits emblazoned with the Virgin logo. The mother ship lifts off and climbs for about 45 minutes to an altitude of 47,000 feet. At the moment of release, the spaceship does not merely drop but pitches down for a clear separation. The maneuver is felt by the passengers as a slight negative G that raises them against their seat belts—the same sinking sensation that can sometimes be felt on a commercial jet in moderate turbulence. Then the co-pilot ignites the rocket, and there is nothing moderate about it. The motor achieves full thrust within a second and shoves the aircraft forward with a relentless three-G acceleration that pushes the passengers against their seatbacks and keeps them there. The G-load in that axis is relatively benign, because it does not drain blood from the head, but, in combination with the vibrations and noise, it may disorient some passengers and will likely surprise most.

Accelerating through Mach 0.95, the aircraft wobbles as shock waves develop on its wings and tails. This is known as a burble, and it marks the entry into supersonic flight. The shock waves change the airflow over the conventional control surfaces—the elevons—and render them almost useless, forcing the pilot in supersonic flight to fly entirely by trimming the stabilizers on the tail. Flying by trim is difficult to do well, but with pilots like these the passengers probably don’t need to worry. At Mach 1, the pilot rolls the pitch trim aft to a pre-determined position, and the spaceship responds by bending the flight path upward at a rate that pushes the passengers straight down into their seats with a force of 2.5 Gs. The passengers are now experiencing a total of 5.5 Gs, divided between two distinct vectors, and are rotated onto their backs as the spaceship accelerates ever more steeply upward. As they approach the vertical, nearing Mach 2, the pilot rolls the trim forward to capture the position, and 2.5 Gs are stripped away. Pointing straight up, the ship rockets into air growing so thin that the aerodynamic speeds decrease rapidly even as the ship keeps accelerating through Mach 3. At around that time, after about one minute of burn, and when an onboard instrument shows that the vehicle has sufficient energy to follow a ballistic path into space, the pilots shut down the rocket motor. The effect for the passengers, who are lying on their backs, facing straight up, is to go immediately from a condition of three Gs to the zero-G state called weightlessness.

This has little to do with being in space. In fact, Branson’s passengers, now at an altitude of about 150,000 feet, are only halfway there. But even if they were all the way there, somewhere above the Kármán line, or three times higher, where the shuttle flew, it would not mean that the earth’s gravitational pull had somehow been escaped. Indeed, the force of gravity at those altitudes is nearly the same as on the ground. Objects achieve low orbits not by levitation but by the energy invested in their speed: with no atmosphere to slow them down they travel so fast horizontally (at least 17,500 miles per hour) that, as they drop toward the surface, their path matches the curvature of the earth. They go into an undiminishing free fall around the planet. It is the free fall, the vertical acceleration, that produces zero Gs. This may be intuitively obvious when it comes to the initial descent from the apogee, before the atmosphere begins to slow things down, but it is equally true during the ascent, after the rocket motor cuts out and the vertical deceleration is due purely to gravity’s pull, without the complications of aerodynamic drag. Pure accelerations, negative or positive, have the same effect. Slowing while going up feels exactly like falling down.

An abrupt transition from three to zero Gs is what shuttle crews went through. A shuttle commander told me it was disconcerting the first time, like stepping off a cliff into a vacuum. It helped to remain strapped into a seat for a while, and then to avoid unusual movements for the first day or two. But aboard the Virgin spaceship, this is the moment the passengers have been waiting for, and they won’t have time to get acclimated. After a pilot gives the O.K., they can push a single release button, free themselves from their straps, and go floating around. They have about four minutes of this before needing to settle down. Some may be so stunned by the rocket flight that they don’t dare release, but we can assume that having gone this far most will follow through with the plan. Once they release, they can have a great time: do somersaults in midair, assume yoga positions, think lofty thoughts about life on earth, and try not to kick one another in the face. We have also entered here into the realm of projectile vomiting. Virgin Galactic insists that this will not be a problem, but presumably it will equip passengers with quick-access sick bags. Whatever happens in the cabin, the passengers are on their own; the pilots remain strapped in and cannot move aft to help.

Investigators examine the wreckage.

From Xinhua/Polaris.

Though slowing, the spaceship is still climbing rapidly. A few seconds after the rocket shutdown, as the vehicle shoots through 210,000 feet, the air outside becomes so thin that the aerodynamic forces amount to the merest breeze. The pilots activate the reaction-control system—the compressed-air jets—and raise the tail into the feathered position. The action of feathering at this point uses the last of the outside air to gently pitch the spaceship onto its back. Very soon after that there is essentially no drag on the vehicle; it is above the atmosphere, though not yet in space. It keeps climbing and would do so in exactly the same manner, no matter what the attitude—tail first, sideways, whatever—but through the overhead windows the upside-down position gives the passengers the best view of Earth below. In this attitude, inverted, the spaceship climbs through the 100-kilometer Kármán line and then goes 10 kilometers higher, to an apogee of 361,000 feet—at least if it performs according to the original plans. The view from here stretches for hundreds of miles and includes gently curved horizons topped by a thin line of atmosphere, with the blackness of outer space beyond.

From its apogee the ship now begins to fall, accelerating vertically to more than Mach 3 and a rate of descent of 200,000 feet per minute. During the fall, the pilot uses the reaction-control jets to pitch the aircraft right side up again, with its wings level and nose slightly below the horizon—the attitude it will be forced into by the feather as it descends into the upper atmosphere. The procedure is not necessary for safety but helps to avoid the violent pendulum oscillations that Mike Melvill endured. By now the passenger seats have been reconfigured to a reclined position in anticipation of the G-loads to come. The passengers are still floating around. Virgin Galactic has run zero-G tests with a cabin mock-up being flown in brief parabolic arcs, and believes that, when the pilots give the word, ordinary passengers should be able to return to their seats and strap in. This remains to be seen. Those who do not succeed in reaching their seats will have to lie in the aisle for the re-entry. The G-loads come on gently at first, but then ramp up quickly as aerodynamic drag slows the descent. The total time experiencing loads above 2 Gs is about 40 seconds. The loads build to 5.5 Gs, then subside. This is not a dual-directional force but a single vector toward the cabin floor. The pilots are conditioned to take it sitting up. Because of their reclined positions, few if any of the passengers will have trouble.

By the time the re-entry ends, at about 60,000 feet, the descent rate has slowed to subsonic speeds. The pilots unfeather the tail, returning it to a conventional position, flush with the fuselage, and locking it there. This pitches the nose down steeply, about 45 degrees below the horizon, and puts the aircraft into a dive. The pilot pulls out of the dive and enters a relatively sedate glide at 185 miles an hour. The passenger seats are returned to the upright position. The glide lasts 20 minutes and finishes with a touchdown on the runway. It is not clear how many of the passengers will be in a condition quite yet to celebrate. But it is doubtful that any of them, having invested so much in the experience, will regret having made the flight. All that is in the future. But, barring another accident, it is a future that seems certain to come.