On several occasions after the breakup of the Space Shuttle Columbia on February 1st, I found myself in conversations about the safety of the space program. Actually, my knowledge is limited to much lower altitudes and speeds and much more prosaic equipment. On a few occasions, nevertheless, I felt obliged to put in my two cents when someone commented that the Shuttle was obviously unsafe, given its record of one fatality for every 10 or so flights.
It's hard to define an expected level of safety for something like the Space Shuttle. Considering the extreme difficulty of what it does, it has succeeded very well, though, like Concorde, it has been an economic failure. The appearance of safety depends a lot on the numbers that you select to represent it. The one-in-10 figure sounds awful; if you say one accident per 60 launches it sounds a bit better. But try putting it in terms of passenger miles. A week-long Shuttle flight with seven aboard racks up some 20 million passenger miles, and if you think of 1.2 billion passenger miles per disaster, it starts to sound almost good. That's the equivalent of 4,000 transcontinental flights of a half-empty 757.
NASA's jolly publicity photos of smiling orange-suited astronauts (what a lovely name-star sailors!) seem intended to suggest that space flight is a walk in the park. Well, maybe Central Park. It's risky, and it's going to remain risky for a long time. The astronauts know it. The technology of the Shuttle is comparatively ancient, but various projects for replacements have come and gone, including ones that dispense with the now-famous insulating tiles, while the old Shuttle soldiers on.
With an ever-shrinking budget, NASA has attempted to shift some of the burden of space vehicle development to private industry. Shuttle operations are now managed by a private contractor, a subsidiary of Boeing. Lockheed-Martin benefited from a period of NASA enthusiasm for a proposed single-stage-to-orbit vehicle called VentureStar, a huge lifting body that was supposed to launch vertically into orbit, glide back, and be ready for reuse in the twinkling of an eye, unlike the Shuttle, which requires the ministrations of armies of technicians for weeks or months between flights. A fundamental part of the VentureStar concept was its so-called linear aerospike rocket engines. An ingenious departure from the familiar mushroom-shaped design, the linear aerospike looks a little bit like an automotive V-8. Rows of combustors eject hot gas against two curved surfaces that form an aft-pointing vee. The outer sides of the exhaust plume are free to adjust their shape to ambient pressure and velocity. Despite high expectations of increased efficiency and many promising ground tests, the linear aerospike had not reached flight status when NASA, with the fiscal belt, or noose, growing ever tighter, cancelled the program.
Another abortive NASA/industry project was the X-34. An unmanned booster system for small to medium-sized payloads, it was essentially a scaled-up version of Orbital Sciences Corporation's successful Pegasus, a rocket-powered spaceplane launched from under the wing of a high-flying L-1011. The business end of the partnership foundered because one of the participants, Rockwell, refused to use the most suitable engine available, which happened to be of Russian manufacture. Rockwell wanted to use its own brand, though it was less powerful and would have made for a less capable booster. Eventually, citing irreconcilable differences, Rockwell withdrew from the program in 1996. The remaining partner, Orbital Sciences, decided to go it alone under a new NASA contract, but that one was cancelled as well for the usual reason-money. Such are the vicissitudes and ignominies of privatizing the Last Frontier.
