So what happened in the skies above Mojave, California, yesterday morning when Virgin Galactic’s SpaceshipTwo exploded and crashed?
The press conference at 2 PM was sparse with information, not even releasing the names of the pilots who were killed and injured, despite family having been notified. Based on eyewitness accounts (one of which was apparently viewed through a telephoto lens), SpaceShipTwo, the vehicle designed to take tourists and researchers into space, started up its rocket engine in a test flight, and a couple seconds later, the engine exploded. The explosion seems to have killed the co-pilot, who went down with the remains of the aircraft, and the pilot bailed out. He was reportedly injured, either in the explosion, or perhaps on a hard parachute landing, and taken to Antelope Valley Hospital thirty miles south, in Lancaster.
What caused the accident? I’d like to say I’m surprised by it, but I’m not, really. Many in the industry, including me, have been concerned about Virgin’s propulsion system for years. Ironically, it was chosen for its ostensible safety characteristics. The theory was that a hybrid rocket engine (liquid oxidizer, solid fuel) couldn’t explode, because the solid fuel couldn’t mix with the oxidizer, and the burning could only occur on the surface of the fuel slug. But about a quarter of a century ago, at the rocket lab a few miles east of today’s crash site, a large LOX/rubber hybrid engine launched itself from its test stand down a mountain in a test in which the nozzle clogged, pressure built up, and the end cap blew off. Anyone present knew that the notion that hybrid rockets were “safe” was nonsense.
More to the point, a previous hybrid engine design for SpaceShipTwo itself had, in fact, exploded on a test stand in July of 2007, killing three Scaled Composites employees and injuring several others, without even having any fuel present. The company has never released details on what the root cause was, but nitrous oxide (aka the “laughing gas” in the dentist’s office) can, under certain conditions, become what is called a “monopropellant.” That is, it can burn all by itself, and that is probably what caused the fatal explosion over seven years ago. We don’t know what happened yesterday, but the eyewitness with the high-power camera said that it looked as though the oxidizer tank exploded.
Doug Messier, long-time Mojave resident and proprietor of the Parabolic Arc blog, published a post on Thursday night that today seems prescient, and provides some history and context for Friday’s tragedy (and it is a tragedy, in the classical sense, perhaps complete with hubris):
In terms of its propulsion system, SpaceShipOne was actually a step backward. The X-15 had used the XLR-99, a sophisticated bi-propellant liquid engine that could be throttled, restarted and used multiple times. It was complicated and prone to failure; one blew up on Scott Crossfield during a static test, destroying the vehicle but sparing the pilot’s life.
Rutan steered away from liquid engines; he viewed them as being overly complicated and possessing too many failure modes. Instead, he developed a novel hybrid motor that used nitrous oxide (laughing gas) to burn a large chunk of rubber fuel. SpaceShipOne was the first time a hybrid engine had been used in human spaceflight.
The hybrid worked well enough for SpaceShipOne. However, the motor ran rough, shaking the ship due to the uneven burning of the rubber. On one flight, the pilot heard a loud bang and feared the ship’s tail had been blown off. It turned out to be a chunk of rubber that had shot out the nozzle. The tail was still there.
The hybrid also was expensive because the rocket casing containing the rubber and the attached nozzle needed to be replaced after each flight. Like the space shuttle, the partially reusable nature of SpaceShipOne drove up operating costs and complexity. It was like driving a car from Mojave to Los Angeles and back, and then installing a new engine before making the trip again.
After the Ansari X Prize, some people tried to convince Rutan to replace the hybrid with a reusable liquid engine. He rejected the advice. Rutan came out of SpaceShipOne’s short flight test program believing the hybrid engine was simple and safe, and that it could be easily scaled up for the much larger SpaceShipTwo. He was wrong on both counts.
Everyone is talking about the “new fuel” in the engine today, thinking that this might be the cause. The original design burned rubber; this one uses nylon. The engine had been ground tested, but as Doug notes, it had had problems. This was the first flight test of it. But, while this is speculation only, and we won’t know until the NTSB shows up and they conclude an investigation, I suspect, as noted above, that the analysis will show that the fuel wasn’t really the issue; the oxidizer was.
I also suspect that this will be the nail in the coffin for the hybrid program. Many in the industry (including me) told Virgin after the fatal test-stand explosion in 2007 that they needed to redesign and go to a liquid engine, but they had designed the airframe for the hybrid, and a liquid would have had different mass characteristics, probably necessitating a spacecraft redesign, which they were loath to do. The company appears to have been stuck in a “sunk cost” trap for years, with the belief that, despite all the years of delays, just a little more money and a little more time would finally get them to a workable engine. Yesterday morning, the trap snapped shut.
It’s not clear what their path forward will be, even after the investigation is complete, but they probably will have hard decisions to make. This is particularly bad news for New Mexico’s Spaceport America, which has been waiting for years for Virgin to start flying, and the state taxpayers’ investment to show a return.
Does this mean that the FAA should have been more involved in regulating this activity? No. This was a test flight. This is what test flights are for. The desert around Mojave has a decades-long history of smoking holes in the ground and test-pilot funerals, though most of them usually originate from Edwards AFB, a few mile to the south. The company won’t be flying passengers until they’ve had many successful consecutive test flights, with whatever new vehicle they develop, assuming they follow through on pledges at the press conference yesterday, and move forward.
But while there has been a lot of talk today about what a setback this is for “space tourism,” that implies that Virgin Galactic, the space ship’s owner, was synonymous with that industry. Simply put, it is not, despite the company’s ongoing hype. While it was a huge setback for Virgin Galactic (and Spaceport America), XCOR Aerospace is assembling their own Lynx space plane in a hangar on the Mojave flight line next to the one where the doomed vehicle had been stored. Unlike VG, they took a more traditional approach, first perfecting an engine design, and then designing a vehicle around it. The company expects to have it in flight test next year, with hopes for commercial flights later in the year.
It may be that the number of people who will fly in space in the near term will be reduced, because SpaceShipTwo could fly several passengers at a time, while Lynx will only fly one. On the other hand, it wasn’t clear how fast the former would be able to turn around, with its hybrid rocket engine, whereas Lynx is designed to “gas and go,” and be capable of several flights a day. In addition, we don’t know what Blue Origin is up to, but they have their own plans for suborbital spaceflight (and orbital as well), and no shortage of money if Jeff Bezos is willing to spend it. So space tourism will move forward regardless of what Virgin Galactic does.
But what both yesterday’s accident and Tuesday’s (completely unrelated) Antares disaster at Wallops show is not that “commercial” spaceflight doesn’t work, but that all spaceflight is still risky, and that we need multiple paths and providers. Let’s hope that Congress doesn’t use these events as an excuse to attempt to return it to a government monopoly.