How SpaceShipTwo was designed to work and what happened instead

Efforts to overcome barriers can be hard and fraught with setbacks before the barrier ultimately topples.

That is one persistent message coming out of the breakup of Scaled Composites' SpaceShipTwo over the Mojave Desert last Friday, a tragedy that killed the craft's co-pilot and seriously injured the pilot.

The prototype craft – with an unusual design for reentry that appears to have contributed to the accident – aims to break an economic barrier. With it, Sir Richard Branson's spaceflight company, Virgin Galactic, aspires to providing affordable, regular access to suborbital space for tourism and research.

Yet the goal of having suborbital space as a regular, if brief, destination for people is itself novel – a task in which Scaled Composites, as well as Blue Origin and XCOR Aerospace, are serving as pioneers.

Since the 1960s, astronauts, cosmonauts, and test pilots certainly have reached suborbital space – meaning their spacecraft carries them into space but then quickly arcs back down to Earth. But this usually was a steppingstone toward orbital flight, not an end in itself.

Humans have gone suborbital only eight times, one of which was the result of a launch accident: two US Mercury missions in 1961; two US X-15 flights from Edwards Air Force Base in 1963; an aborted Soviet Soyuz launch in 1975; and three flights of Scaled Composites' SpaceShipOne in 2004, which earned the company the $10 million Ansari X Prize.

As a result, “the operational experience for manned suborbital spaceflight is very limited,” noted an analysis published in 2011 by the Aerospace Medical Association's Commercial Space Flight Working Group.

“These folks are out there doing something really new and different,” says Kerri Cahoy, an assistant professor of aeronautics and astronautics at the Massachusetts Institute of Technology in Cambridge, Mass. “They're developing technologies and capabilities we haven't had.”

Although many of the design needs for a suborbital craft to carry people are simpler than for orbital craft, this still represents new design territory, she suggests.

Of the three suborbital companies, Scaled Composites is on the leading edge, a point indirectly acknowledged by Christopher Hart, acting chairman of the National Transportation Safety Board during an update Sunday evening on the agency's accident investigation.

“These are unique machines,” he said.

But Scaled Composites also is “going through the do-this, don't-do-that development phase,” Dr. Cahoy adds, a phase in which one goal is to refine flight procedures and other key aspects of operation, as well as to improve the craft's performance.

The NTSB is looking at a broad range of issues to pin down the root cause of Friday's tragedy. In addition to hardware, investigators will look at the company's training system, safety culture, and design procedures. They'll also probe for any pressure to hew to a test or final rollout schedule, as well as other issues.

Initial speculation outside the NTSB centered on SpaceShipTwo's rocket motor, which was using a new type of solid fuel for the first time in flight.

The modified motor had been tested on the ground successfully several times, according to Scaled Composites officials. And when NTSB investigators found the motor as well as the craft's fuel and oxidizer tanks among the wreckage, they all were intact, Mr. Hart said. None showed signs of burn-through or breaching.

Telemetry and videos of the crew, however, indicated that the craft's unique reentry system had activated at a particularly inopportune time. Hardware designed to slow and stabilize the craft moved into its reentry configuration nine seconds after the rocket's motor ignited. Coming at a time when the craft was accelerating, the action would have put significant stress on the craft's structure.

The approach that the winged SpaceShipTwo uses is known as feathering. It's a concept that Dean Chapman, an aerospace engineer for NASA's predecessor, the National Advisory Committee for Aeronautics, described in a technical paper in 1958 as part of his overall effort to refine formulas that engineers could use to design reentry systems for spacecraft.

On reentry, where the atmosphere is thinnest, a device would increase a winged craft's drag, slowing it and reducing the heating that the craft experiences, while keeping it stable. Once the craft slowed sufficiently, the device would retract or become part of the wings, and the craft would glide to its landing site.

Because a suborbital vehicle reenters at a far slower pace than a craft reentering from orbit, the approach is said to be best suited to suborbital craft.

SpaceShipTwo achieves the effect with an aerodynamic boom, or feather, at the end of each wing. Each feather extends aft to end in a tail that sports a horizontal stabilizer. During launch, the feathers are set so that the tails and stabilizers are in a normal flight position.

During the initial phase of reentry, however, the crew activates the feathers. The front ends tip down, and at the rear the tails and stabilizers rise up so the stabilizers can act as air brakes. In addition, the feathers keep the craft stable. Once the craft slows sufficiently, the feathers return to their normal flight position for gliding and landing.

For SpaceShipTwo, activating the booms takes two steps involving two separate controls, the NTSB's Hart explained. The crew must first unlock the mechanisms that rotate the booms. Then they must activate them to lift the feathers into the reentry position.

The system worked well for SpaceShipOne, as well as on previous test flights for SpaceShipTwo – up to now.

Telemetry and video showed that SpaceShipTwo had a normal launch, with the engine igniting at the appropriate time. Nine seconds after the motor ignited, telemetry indicated that the feather mechanisms had been unlocked, with video showing the co-pilot unlocking them. Seconds later, telemetry showed that the feathers were being deployed, even though the video showed no evidence that the crew had taken that second step. Then the video went dark and the telemetry feed from the craft stopped.

The craft was hurtling along at just over the speed of sound, or just over Mach 1, and would have been accelerating when the feathers deployed, a process that wasn't to have begun until the craft reached Mach 1.4, Hart said.

“This is a statement of fact, not a statement of cause,” he emphasized, noting that the team still has a great deal of evidence to sort through and eyewitness interviews to conduct.

“The investigation is far from over,” he said.