Antares and SpaceShipTwo: A Wakeup Call for Lots More Test Flights

I remember reading back in 2008 that Virgin Galactic would soon begin flight testing its SpaceShipTwo spaceplane. I recall reading that there would be at least 50 test flights before there would be any thought of the vehicle carrying passengers, other than its two test pilots. I thought, "Hmm, 50 flights... that seems reasonable." I figured that should be enough to work out any kinks and establish the vehicle's relative safety before offering up its services to the public.

Over the next six years, Virgin Galactic did indeed conduct more than 50 test flights -- 54, to be precise. However, only four of those flights (including the failed one on October 31, 2014) were powered flights in which the SpaceShipTwo flew of its own accord. The other flights involved takeoffs of its White Knight Two carrier aircraft with the SpaceShipTwo attached to it. The White Knight Two would fly to an altitude of about 50,000 feet and then release SpaceShipTwo, after which the spaceplane would glide down and land on a runway.

In other words, in all but four of the test flights, SpaceShipTwo's rocket engine was not turned on and used for powered flight.

I think this is a crucial point, because it goes to the heart of whether a rocket -- be it a reusable spaceplane bound for Earth sub-orbit or a more traditional orbital expendable launch vehicle like Orbital Sciences' Antares, which blew up seconds after launch on October 28 -- can realistically be declared "ready" for operational flights after only a handful of test flights (powered ones, that is). Note that the Antares, which failed on its third operational flight, had logged only one test flight and one demonstration flight before it began launching cargo for NASA to the International Space Station (ISS).

Compare this to the 199 test flights (all but one unpowered) that NASA conducted of its X-15 rocket plane between 1959 and 1968. Like SpaceShipTwo, the X-15 was also piloted and air-launched, but by a U.S. Air Force B-52 aircraft. Also compare it with the nearly two dozen test launches conducted between 1961 and 1968 with Little Joe and Saturn rockets in preparation for the manned Apollo missions.

Granted, the comparison with the X-15 may not be entirely fair, because the ultimate goal was not perfecting the vehicle but testing technologies that could be used to eventually build a full-fledged piloted spaceplane. But the point is that actual flight testing of rockets or rocket-powered vehicles tended to be more of a long-term thing back in the 1960s than it is today.

Perhaps we feel like we've already gone through the most difficult part of the learning curve when it comes to rocketry, and that now it's just a matter of perfecting a few minor technical issues. Judging by last week's twin failures, I'm not so sure that's true.