Shirt-Button Turbines

Look inside a laptop computer and you’ll probably find that the largest and heaviest single component is the battery. Despite the remarkable progress scientists and engineers have made in shrinking the electronic and mechanical components inside portable PCs, battery technology has obstinately refused miniaturization. Now a research team at MIT’s Gas Turbine Laboratory has embarked on an ambitious project to develop a turbine engine the size of a shirt button, linked to a tiny electric generator and capable of producing 10 to 20 times as much power as the best chemical batteries.

Spurred by a wave of interest in MEMS devices-microelectrical and mechanical systems fabricated like silicon chips-lab director Alan Epstein and his colleagues began investigating miniature turbines about four years ago. MIT’s Lincoln Laboratory in Lexington, Mass., provided some seed money to start research on the topic, and the project eventually drew the interest of the U.S. Army Research Office, which in 1994 agreed to provide the group $5 to $6 million over five years in the hope that microturbines will provide portable power sources for individual soldiers.

Like a conventional jet engine, the miniature turbine will include three key components: a combustion chamber, a turbine wheel, and a compressor wheel. Fuel burning in the combustion chamber sends exhaust gases through the blades of the turbine wheel, causing it to rotate, which in turn drives the compressor wheel via a central shaft. The vaned compressor rotor then draws outside air into the chamber to feed the burning of more fuel.

One of the first questions the researchers sought to address was whether turbine technology could generate sufficient power at such a tiny scale. “The key to achieving high power density in rotating machinery is high peripheral speed,” Epstein explains. Peripheral speed is the speed at which the outer rim of the turbine’s vaned wheels turn. Spinning at 2.5 million revolutions per minute, the tiny wheels need to reach a peripheral speed of 300 to 600 meters per second-perhaps twice as fast as conventional turbine rotors. High peripheral speed in turn implies high peripheral stress: if the rotor’s materials are not strong enough to withstand this stress, they will fracture.

Members of Epstein’s team conducted a two-year scaling study to investigate the prospects for miniaturization. To their surprise, they found that many of the obstacles they anticipated did not materialize-and that miniaturization actually bestowed some advantages. “We felt initially that scaling effects were going to cause problems relating to the viscosity of the air,” Epstein says. Viscous forces in the air are larger at microscales because air molecules are proportionally larger. A butterfly, for example, must expend proportionately greater energy overcoming air viscosity than a Boeing 747.

As it turned out, microscale air-viscosity effects were not great enough to require a substantially different design approach. In addition, team members realized that since microscale materials are likely to have fewer flaws than macroscale materials, they are proportionately stronger. As a result, the tiny rotors can actually be spun at higher speeds with a lower risk of fracture than conventional turbine rotors.

Armed with this information, the team began to develop and test the turbine’s components. Last spring the researchers demonstrated the feasibility of a 2-millimeter-long combustion chamber and constructed a turbine wheel 4 millimeters in diameter out of silicon, using micro-fabrication techniques similar to those used to mass produce computer chips. They are now testing the low-friction air bearings that will permit the turbine wheel to spin at full speed. According to Epstein, the design of the third major component in the engine, the compressor wheel, will require only minor changes in the turbine wheel.

Over the next few months, the researchers plan to demonstrate the viability of a thin-film electric starter-generator, to be mounted on a shroud over the compressor blades, and then embark on the next stage of their project: integrating the components onto a single silicon chip. Since the high-temperature performance of silicon is limited, the team is simultaneously investigating methods for manufacturing microturbines out of silicon carbide, a tougher, more temperature-resistant material.

Ultimately, the researchers hope to produce a prototype turbine power plant weighing less than one gram and generating 10 to 20 watts of electricity. If this high-risk effort proves successful, inexpensive miniature turbine generators could eventually be stamped out in large quantities like computer chips, Epstein says, and microturbines could become as ubiquitous in civilian life as batteries.

“It’s a very difficult technology development problem,” says James H. Smith, micromachine technology team leader at the U.S. Department of Energy’s Sandia National Laboratory in Albuquerque, N.Mex., “but the MIT group knows more about this than anyone else and they are making excellent progress. If they can solve the problems, the impact would be tremendous.”