Peewee Power

The invention of a gas-fueled generator the size of a quarter heralds a future of ever-smaller machines

The lightbulb went on when Alan Epstein, a meticulous 53-year-old engineer with a twinkling sense of humor, was sitting in a jury pool in a Cambridge, Massachusetts, courthouse. The question he was pondering that spring day in 1994 had been put to him by a United States Air Force official: How difficult would it be to build a jet engine the size of, say, a coffee cup? Epstein, the head of MIT’s Gas Turbine Laboratory, which has a staff of 80 and receives $8 million in research funds annually, had answered: difficult.

But as he awaited his jury assignment, he reconsidered the question. What if a tiny turbine could be built using the approach perfected by computer-chip makers? That is, construct it not out of thousands of steel, aluminum and copper parts welded and riveted together, but etch it instead on a wafer of silicon. A turbine made that way, he thought to himself, could even be as small as a shirt-collar button.

To understand how Epstein’s colleagues at first reacted to his notion, keep in mind that although turbine engineering has many satisfactions, the mischievous pursuit of startling innovations is not usually one of them. Advances in the field tend to be incremental in the extreme. Because turbine failure can be so catastrophic—the engines propel jet aircraft, after all, and are used by electric utilities to generate power—caution is the rule in the esoteric world of turbine R&D.

The way Epstein remembers it, every time he presented the idea of a microturbine etched out of silicon to his colleagues, they laughed in his face. But nobody’s laughing now: Epstein and his coworkers have won over doubters with a prototype turbine the size of a quarter that, though crude in some respects, will eventually generate enough power to run a cell phone. That may not sound like much to show for seven years of high-intensity research, but turbine experts say it’s a huge accomplishment. Commercial and military applications for such a power source abound. Ounce for ounce, a minuscule turbine generator fueled by kerosene could produce 10 to 20 times the power of a conventional battery. A computer laptop might run continuously for a week on one cartridge; a cell phone cartridge might fuel 72 hours of conversation. Ultimately, though, Epstein speculates, the virtue of microturbine-powered consumer electronics won’t be how much longer they operate but how much smaller and lighter the components might be.

A chief supporter of the microturbine generator project is the U.S. Army, which wants portable, lightweight power sources to run an arsenal of electronic devices such as radios, computers and satellite-based navigation equipment. "As soldiers get more electronic and they’re farther from outlets, power sources become more important," Epstein says.

In Epstein’s view, machines are in the midst of a design revolution akin to what electronics went through decades ago, when transistors and computer chips spurred development of smaller and smaller products. "The idea of making micro devices is spreading very rapidly throughout every discipline of science and engineering," he says. And his team’s contribution to the emerging field of what is known as microelectromechanical systems, or MEMs, is the microgenerator that his lab has produced. "I think we invented a new field," Epstein says. "Power MEMs."

It was no small feat squeezing the concept of the huge roaring turbine engine that we’re all familiar with into a quietly whirring device the size of a coin. "There were many times that our brains were fried," says Stuart Jacobson, who worked for the project and is now in private industry. "But you would be hard-pressed to find anyone that would tell you that it wasn’t one of the best experiences they ever had." The research has "changed the way we think about turbines," Epstein says. For instance, it demolished information long presented in textbooks that a turbine combustion chamber, because of a presumably unyielding engineering limit, had to be 18 inches long. The new research proved the received wisdom wrong, showing that the combustion chamber "could actually be scaled down," he says.

Conventional gas turbines use the explosive power of combusted fuel to spin fanlike blades at high speeds; the blades can create thrust, as in a jet engine, or electric current, if attached to a generator. One obstacle to shrinking a turbine is what might be called fan physics. For a gas turbine to work well, the tips of its rotors have to turn at near the speed of sound. But the smaller the turbine, the faster the rotor must spin for its tips to achieve near-sonic speed. Thus, while a conventional jet engine’s rotor spins efficiently at about 20,000 revolutions per minute, the MIT microturbine rotor has to turn more than 100 times faster, or two million rpm—more than 20,000 revolutions per second.

Epstein and his colleagues figured that a shaft turning at such an extraordinary speed would quickly wear out. So the 50-member team came up with a way for the rotor to use its blistering speed to levitate, hovering precisely in place like a miniature, supercharged Frisbee. Instead of ball bearings, it has "air bearings," Epstein says.

Frank Marble, retired Cal Tech professor and eminence of turbine engineering, thinks Epstein’s tiny engines have a bright, if yet unknown, future. Like the jet engine, whose use on civilian aircraft was a revolutionary afterthought, "we probably will use these microturbines the most in areas we conceive of the least," says Marble.

Epstein is the first to caution that the technology still isn’t ready for prime time. The microturbine generates a lot of heat, and he has yet to implement a solution for cooling the mechanism or its exhaust. After all, no one’s going to put a scorching hot cell phone to her ear. For now, Epstein says, "we have the world’s first jet-powered hair dryer." That’s no mere joke. He has actually been approached by a company interested in producing a hair dryer that isn’t plugged in (always an electrocution hazard). As it is, Epstein’s group has a contract with the Army that calls for the scientists to demonstrate a microturbine generator within four years. That deadline, he says, is no problem.

The development of the microturbine, which has borrowed heavily from the information revolution’s knack for chip making and for scaling big ideas down to size, augurs a sci-fi future of incredible shrinking machines, from artificial organs to pipsqueak robots, according to the experts. "The future is small," Epstein says. In other words, micromachines could well be the Next Big Thing.

Get the latest Science stories in your inbox.