Interplanetary missions can burn as much as 50,000 pounds of fuel to get a 1,000-pound probe out of Earth's orbit. While many researchers have labored to reduce the staggering inefficiency of rockets, a nuclear engineer from the University of Illinois at Urbana-Champaign has devised a completely different booster scheme, one that involves no onboard fuel. Clifford Singer proposes a system in which a string of satellites, each separated by only a mile or so, toss tiny ball bearing*#151; size pellets at a space probe. The momentum of the pellets would propel the probe.
Singer's idea grew out of research he began some 20 years ago while studying how to launch pellets of frozen hydrogen into a reactor where powerful electric fields would zap them, causing a nuclear fusion reaction. He found that the pellet launcher of choice was a type of centrifuge. Pellets loaded into the spinning centrifuge were flung into the reactor at high speed.
Singer wondered if he could apply a variation of this idea to spacecraft. His scheme, which exists only on paper, calls for a system of small orbiting satellites--each weighing about 20 pounds--that whip BB-size steel pellets back and forth like a demented jai alai team. The satellites, in low-Earth orbit, would each house a one-foot-long spinning arm (see illustration) that would both launch and catch pellets.
The first satellite in the chain shoots a stream of pellets, aimed by an onboard computer, at the next satellite down the line at speeds of about a mile per second. (Singer says that aiming the pellets shouldn't be a problem--without air resistance, trajectories are easy to control.) As the relay satellite catches the pellet, it absorbs the pellet's momentum and gets a small push. The spinning arm on the relay satellite accelerates the pellet and shoots it toward the next relay. As the satellite passes the pellet on, the momentum of the toss pushes the satellite back toward its initial position. A pellet caught and tossed by a moving satellite adds the satellite's speed to its own, so that after a few relays the pellets could speed at several miles a second.
The spacecraft at the end of the line would catch the pellets and whip them back toward the start, not only gaining speed from both catching and tossing, but also returning pellets to the satellites for reuse. Depending on the size of the spacecraft, a few thousand pellets might do the job, and the whole system could run on solar energy. Conventional rockets would still need to lift the satellites into low-Earth orbit, but Singer figures his scheme could cut fuel costs by more than 90 percent.
As few as three relay satellites, Singer estimates, could accelerate a probe to interplanetary speeds. "Six kilometers a second [about 13,000 miles an hour] is the maximum for a planetary mission unless you are in a terrible hurry," he says. "I'm pretty sure we can get to that." He plans to apply for a NASA grant and hopes to have a working tabletop model of his system started by the end of next year.