Flywheels put a new spin on electric vehicles.

Advances in high-strength composite materials, frictionless magnetic bearnings, high-efficiency motor/generators, and lower-cost miniaturized power conditioning and control electronics have resurrected the possibility that the venerable flywheel could be used to power pollution-free electric

ENGINEERS WORKING TO develop "zero-emissions" cars and trucks for the future are facing an extremely tough task in trying to supplant the long-dominant internal combustion engine, which environmentalists tell us wastes energy and pollutes the air. Not only must any replacement drivetrain produce pollution-free power with high efficiency, reliability, safety, and affordability, it must also meet the expectations of drivers accustomed to the high torques and accelerations provided by gasoline- and diesel-fueled internal combustion (IC) engines. Researchers developing altenative vehicular power sources are finding it hard to match the extensive energy content of cheap and readily available hydrocarbon fuels.

Carmakers are acutely aware that the first company to offer an affordable, resonably performing, and environmentally friendly automobile will have a significant advantage over competitors, especially if government environmental regulations continue to be tightened. The state of California, in particular, has provided automakers with a powerful incentive to develop practical alternatives by requiring companies to ensure that 2 percent of their vehicles have zero emissions by 1998.

As has been widely reported, the current focus is on electrochemical batteries. Although chemical bettery-powered electric vehicles (EV) use energy more efficiently than their IC engine--powered counterparts, the limited range and performance of current EVs are caused by the modest energy and power densities of today's batteries, particularly in cold weather. In addition, chemical batteries typically contain hazardous materials that pose disposal and recycling questions. Battery researchers have worked long and hard to overcome these drawbacks in numerous industry- and government-funded programs, but, barring a breakthrough, the realities of electrochemistry make the prospects for major improvements in chamical battery technology unclear.

There is another alternative, however, that has been around since ancient times--the flywheel, a balanced mass spinning around a constant axis that stores energy as rotational kinetic energy. One of humankind's earliest inventions, the flywheel served as the bases of the potter's wheel and the grindstone. Today, flywheels are widely used in IC engines to convert the pulsing bursts of combustion energy produced in the cylinders into smooth continuous power for the drivertrain. Less well known are aerospace flywheel systems that act gyroscopes to help orient orbiting satellites and store energy collected by photovoltaic arrays in their rotation. Flywheel technology is familiar to readers of this magazine, which published articles on the topic in February 1978, June 1988, and December 1990.

Recently, many researchers have become convinced that modern flywheel energy storage (FES) systems--fiber-composite rotors spinning at many thousands of rpm on frictionless magnetic bearings--could drive generators that provide power for efficient nonpolluting EVs. Also known as inertial energy storage devices or electromechanical batteries (EMB), these systems could theoretically rival chemical batteries in terms of power, energy density, cycle life, charge time, operating temperature range, environmental friendliness, and maintenance needs. FES is now considered a viable technology for recovering braking energy (regenerative braking), averaging peak power demands, and storing energy for electric and hybrid vehicles. FES systems are also being developed for stationary applications such as utility load-leveling systems, uninterrupted power supplies, and storage capacity for solar and wind power systems.

"Research and development in flywheel energy storage has advanced this technology to the point where FES-equipped vehicles could be available in the next few years," said a recent technical assessment prepared for the U.S. Department of Energy (DOE) Office of Transportation Technologies, Electric and Hybrid Propulsion division, by Abacus Technology Corp., a technical consulting firm in Chevy Chase, Md. "High specific energy, specific power, and efficiency have been demonstrated at the component level in numerous laboratory experiments," the report continued. "Flywheels have the theoretical potential to equal or better every one of the U.S. Advanced Battery Consortium's battery goals, with the possible exception of purchase price (at least until production volumes increase). Lifetime ownership costs of a flywheel vehicle, however, are expected to be significantly lower than those of a battery-only EV due to the glywheel's greater reliability, longer service life, and higher efficiency.

"Flywheel energy storage is sufficiently advanced to build units for demonstration in electric vehicles," the DOE/Abacus report said. "All components of FSE systems have been successfully used in other applications or tested in laboratories...All that remains is to engineer, build, and test a complete prototype system." The report recommends that the existing components be integrated into "a complete mass producible energy storage system, optimized for handling peak power demands in flywheel/battery-powered hybrid electric vehicles." The report also advocates designing and building "an appropriately optimized flywheel energy storage system capable of replacing the chemical batteries in conventional electric vehicles."

Tantalized by the possibilities, research groups across the country--in academic or governmental institutions, small start-up companies, and large corporations--are working individually and jointly to develop the first generation of modern FSE technology. Among the known players in the field are: Advanced Controls Technology Inc. of Northridge, Calif.; Allied-Signal Inc. of Morristown, N.J.; American Flywheel Systems of Medina, Wash./Honeywell Satellite Systems of Phoenix, Ariz.; Flywheel Energy Systems Inc. of Ottawa, Ont.; Lawrence Livermore Technology Inc. of Latham, N.Y.; Oak Ridge National Laboratory in Oak Ridge, Tenn.; SatCon Technology Corp. of Cambridge, Mass.; Thortek Inc. of Knoxville, Tenn.; Unique Mobility of Golden, Colo.; United Technologies Corp. of Hartford,

Conn.; the University of Maryland in College Park; and U.S. Flywheel Systems Inc. of Laguna Hills, Calif.

AUTOMAKER INTEREST

Car manufacturers around the world are also interested in flywheel batteries, at least for hybrid appliations in which they would be linked to small IC engines or electrochemical battery arrays to provide extra power when required. Although information is hard to come by, the Big Three American automakers have instituted programs to study and perhaps develop flywheel-powered EVs. Last July, for example, Chrysler Corp. of Highland Park, Mich., announced that it is working with aerospace flywheel developer SatCon on "innovative new vehicle drivetrain components," which is widely interpreted to mean flywheels, among other advanced technologies. Meanwhile, Ford Motor Co. of Dearborn, Mich., and General Motors Corp. of Detroit have reportedly contracted with various undisclosed flywheel groups to investigate this fast-developing technology. Asian and European automakers are also pursuing automotive flywheels. Munich, Germany-based BMW AG is reportedly collaborating with United Technologies on FES technology, as is Unique Mobility.

"Ford is activity looking at FES," said Bradford, Bates, the automaker's manager of alternative power source technology. "If you do the mathematics, FES turns out to have many interiguing aspects, especially the high power densities, which could provide advantages in boosting acceleration and in regenerative braking. But it's premature to make a judgment," Bates said. "For us, the flywheel is still just another black box that could store power to drive a car. So far all the results have been in the lab. No one has a practical system yet, so it's not at all clear it would be a cost-effective approach." General Motors officials echoed Bates' sentiments. Chrysler chose not to comment.

Much of this conservatism is based on "some pretty extravagant claims made hy flywheel proponents recently as well as the tough engineering challenges faced by anyone trying to develop the technology," said Ralph Flanagan, president of Flywheel Energy Systems and a professor of mechanical engineering at the University of Ottawa. A 30-year veteran of the field, Flanagan said that "people tend to simplify the difficulties of developing and integrating into an EMB a suitable magnetic bearing system (including backup or touch-down bearings) that can be mass-produced at a reasonable cost." They also tend to downplay problems involved with "the rotor dynamics when moving this very high-speed rotating equipment around. But because of its inherent specific power and power density properties, the flywheel battery can do things that nothing else can do.

"It's akin to building the Model-T," Flanagan said. "We still have a lot to learn about practical flywheels, but with good engineering and enough financial investment, I believe we'll soon see flywheels used for the power-surge application in hybrid vehicles for urban driving and probably in statinary utility systems." Small electromechanical batteries in the 0.5- to 1.5kilowatt-hour (kWh) range could suffice for hybrid electric-drive urban vehicles such as delivery vans and buses, he noted, while fully flywheel-powered vehicles would make use of arrays of modular versions of these smaller units, similar to the current electrochemical battery arrays in today's EVs. A variety of stationary energy storage needs could be handled by larger modules with 25-kWh cpacities.

The investment part of the equation also seems to be moving along. The U.S. Advanced Research Projects Agency (ARPA) has reportedly alloted a portion of the new multiyear $25 million EV/Infrastructure Demonstration program to flywheel research. Money will be distributed among six regional government-indusstrial consortia established to help convert defense-related industries to peacetime endeavors. In addition, the DOE is reportedly planning to funnel an undisclosed amount of money toward flywheel research as part of the $19 million U.S. Hybrid Propulsion Systems Development Program.

FLYWHEEL VEHICLES

In the late 1940s, Swiss researchers built a flywheel-equipped electric-powered trnsit bus called the Gyrobus, which was used in urban passenger service over hilly terrain. Electric motors at "charging" stops spun up a heavy steel flywheel. Energy was added and withdrawn from it through a reversible motor/generator that powered electric drive motors. A regenerative braking system fed energy back into the flywheel. The Gyrobus' main drawbacks were the low specific energy of its cumbersome steel flywheel and the bulky and costly power conversion equipment used to control electrical energy to and from the drivetrain.

The recently proposed government funding is not the first time U.S. federal agencies have tried to push FES technology. For seven years, starting in the mid-1970s, an American coalition of government, industry, and academia spent about $90 million investigating flywheels. These studies looked at both all-electric flywheel-based vehicles and all-mechanical systems where flywheels were used as "power boosters" to increase vehicle efficiency through mechanically based regenerative braking. Researchers were unable to demonstrate high energy and power densities, however, partly due to the limits of available materials. Old-time flywheel hands claimed that these research programs lacked focus. "People developed a lot of bits of a complete flywheel system piecemeal, but nobody made a product out of them," one said.

In the late 1970s, for example, the Energy Department built a hybrid vehicle called ETV-2, which used batteries for primary energy storage and a lightweight composite flywheel mechanically linked to the motor/generator for power surges. Although the flywheel improved acceleration compared to the earlier batteries-only ETV-1, performance was inadequate for normal driving.

There have also been several attempts to build flywheel-augmented uran buses. Probably the most recent was that of Magnet-morot GmbH of Starnburg, Germany. This system used a glass-fiber flywheel rotating at 12,000 rpm in conjunction with a six-cylinder 90-horsepower turbocharged BMW diesel engine. The diesel drove an electric generator whose current was sent either directly to the drive motor or to the FES unit if there was excess energy, which was to be used later to meet peak power demands. The diesel engine, which was sized for cruising, was only one-third the size it would be if it were not assisted by a flywheel. The system was installed in two buses and tested on routes in Munich a couple of years ago, but no commercial product has yet resulted from the project.

FLYWHEEL FUNDAMENTALS

The main figure of merit of any battery intended for vehicular use is its specific energy, that is, watt-hours of recoverable energy per kilogram of battery weight, according to long-time flywheel (and magnetic fusion) researcher Richard F. Post, a senior scientist at Lawrence Livermore National Laboratory and a professor emeritus at the University of California in Davis. Stored energy in a simple rim or hoop flywheel increases linearly with mass and by the square of the rotational speed. In principle, he said, one could store as much energy as one wants simply by spinning the wheel faster, but the flywheel's speed limit is ultimately set by the tensile strength of the material of which the rim is made.

The wheel's tensile strength, according to Post, must be sufficient to withstand the hoop strees resulting from centrifugal forces, which produce large radial dilations that tend to casue the wheel to fly apart. As with energy stored, these forces are proportional to the mass of the rim and increase with the square of the rotation speed. This means that a higher-density flywheel will store more energy than a lower-density wheel of the same size and shape spinning at the same speed, but it also menas that it is subjected to much stronger centrifugal forces. As the forces on the denser wheel approach the limits of its constituent material's tensile strength much more quickly, a less-dense flywheel can rotate faster. Because energy storage rises with the square of rotor speed, but only linearly with mass, lighter is better for a flywheel. This means that a flywheel designer wants the lowest-density strongest material available, and that means advanced fiber composites.

Among the earliest attempts at composite flywheels were U.S. Navy-sponsored wheels using thick-walled cylkinders fabricated by filament-winding epoxy-impregnated glass fibers onto a mandrel, according to Post. In spin tests they failed by delamination at rotation speeds far below those corresponding to the tensile strength of the composite. This delamination came about because of the radial stresses induced in the composite by the centrifugal force field.

The latest high-strength fibers, epitomized by graphite fibers with tensile strengths of 1 million pounds per square inch, will likely be used in first-generation vehicular FES systems, Post said. Later, more advanced units will probably use graphite fibers with strengths approaching 1.5 million psi, which are predicted to be available in volume in a few years. For other than stationary vehicular applications, less-costly fivers such as E-glass may be preferred, although glass fibers are more subject to fatigue than graphite.

"Current FES designs for a 1-kWh rotor can be realistically expected to weigh from 10 to 11 kilograms," Flanagan said. "A typical rotor geometry could be about 10 inches in diameter with a 6-inch axial thickness operating at 80,000 rpm (rotor tip speed about 1110 meters per second)." These rotors would be good for about 100,000 power cycles operating at temperatures ranging from -- 40 [degrees] to 100 [degrees] C.

MAGNETIC BEARINGS

A sign on a wall in Post's office says, "It's the bearings, stupid." The slogan emphasizes what many FES researchers, including Post, believe is "the number 1 critical problem" that stands in the way of flywheel success.

Bearings have long been a major issue in flywheel development because of the difficult demands they must meet, especially in commercially viable mobile applications, said Edward S. Zorzi, a specialist in high-speed rotating machinery and vice president of engineering and technology for American Flywheel Systems. "In general, you need low-friction high-speed bearings with long service life to limit parasitic losses so as to maximize the potential of usable energy in the system. That means magnetic bearings."

Low-loss noncontact magnetic bearings are attractive, Zorzi said, because they are virtually frictionless and wear little with no lubrication. They can operate in vacuum so that windage losses can be avoided, and they can isolate the rotor from vibrations due to road shocks and gyroscopic loads, in addition to helping with noise-reduction and heat-dissipation issues. In recent years, bearing size and costs have dropped dramatically, as has design complexity. It is also thought that magnetically supported rotors can be less precisely balanced, leading to lower production costs.

In a typical active-type magnetic bearing, a shaft is suspended by magnetic forces as it rotates. These forces, along with the loads that act on the shaft, are controlled and balanced by position sensors and electronic feedback circuits working in concert to introduce stabilizing magnetic flux forces through controlling the currents in electromagnetic windings within the bearing assembly. Passive magnetic bearings, a newer type, use powerful rare-earth permanent magnets--neodymium-iron-boron and samarium-cobalt-and clever geometry to support and stabilize a spinning shaft without resorting to feedback control. Newly developed superconducting materials may make passive magnetic bearings even more practical in the relatively near future.

"Passive bearings would be nice to minimize parasitic losses," Zorzi noted. "On other hand, you would like the magnetic bearing to be active enough to compensate for road shocks and rotor balance to avoid secondary damage." Several researchers noted that some form of mechanical bearing will be needed for backup, in case power to the magnetic bearing fails.

IN A VACUUM

High-performance flywheel systems will have to operate in a vacuum to minimize windage losses, aerodynamic heating, and rotor instability, Flanagan said. They therefore will require leak-proof containment vessels evacuated to about [10.sup.-3] to [10.sup.-5] torr. Passive vacuum systems with "getter" allows [which chemically fix (absorb) unwanted gaseous contaminants] to maintain evacuated conditions would, of course, be preferred over pump-dependent vacuum systems to cut weight in vehicular applications. Although lightweight composite structures would seem to be favored for this application, several flywheel researchers said that metal sheet or metal/composite designs would probably be used because composite vessels lack sufficient vacuum integrity.

The flywheel containment vessel must also serve as a safety enclosure to protect passengers and the rest of the vehicle from rotor rupture resulting from failure of quality control or an accident. Flywheel researchers said that well-engineered multiple-barrier containment structures should be sufficient to hold in any failed rotor debris.

Another major component of a high-performance FES system is the motor/generator (motor/alternator), which spins the flywheel up to speed to store energy and extracts energy by coupling to the spinning rotor to generate electricity for the induction drive motors. In today's designs, permanent magnets spinning with the flywheel rotate past a coil to generate electricity. Alternatively, current can be applied to the coil to cause the magnets on the flywheel to "push" it into rotation.

State-of-the-art electronically commutated rare-earth permanent magnet motor/generators are becoming increasingly compact, efficient, and powerful, said Bruce E. Swartout, another flywheel pioneer and chairman of U.S. Flywheel Systems. "We're talking about a multikilowatt-hour mechanical battery with a motor/generator the size of a roll of LifeSavers," Swartout said. In addition, motor/generator prices are expected to drop with bigger production volumes.

"Sizing motor/generators for different performance criteria is a difficult issue," Zori said. The optimal sizing for cruising down a highway is different from the best choices for acceleration or regenerative braking; consequently you have a tough trade-off involving performance during different phases of operation, he explained. Researchers are said to be considering several different motor/generator configurations to meet this challenge.

Ensuring that rapid power transfers to and from the rotor do not place undue stress on the rotor, bearings, and motor/generator is another key design requirement. In addition, the motor/generator must be kept cool by a small, lightweight, and inexpensive cooling system.

Power conditioning electronics and controls are important FES subsystems. Flywheel motor/generators deliver variable frequency current that, to be useful, must be converted to a lower frequency, with its voltage and phase suited to the load or charging system. Both the active magnetic bearings and the motor/generator will need advanced electronic controls.

Fortuitously, solid-state power conditioning electronics using low-cost metal oxide semiconductor field effect transistors and insulated gate bipolar transistors have been getting smaller and cheaper in recent years, making them suitable for mobile flywheel applications.

ROAD SHOCK AND GYROSCOPIC FORCES

Two other concerns for designers of vehicular FES systems are how to deal with dynamic loads from road shocks and how to handle gyroscopic forces encountered when a spinning rotor changes orientation in maneuvers. In general, transitory loads will be isolated from the rotor by shock-absorbing or elastomeric systems.

The gyroscopic effect is dealt with more easily with small flywheel modules, Post said, because the scaling of the rotor's angular momentum is proportional to the fourth power of its radius, meaning that small rotors will exert markedly smaller gyroscopic effects than larger ones. "The gyroscopic moment of a 1-kWh flywheel at maximum speed is comparable to that of a typical auto engine flywheel at maximum rotation," he said. Zorzi concurred. "You must pay attention to the gyroscopic effect in maneuvering conditions because it is significant, but it's not disabling," he said.

Under consideration are gimbal mountings that allow the rotor to maintain orientation as the vehicle moves, conterrotating rotors that inherently counteract gyroscopic forces, or the use of an even number of FES units, which would ameliorate much of the gyroscopic effect. It should be noted that whatever scheme is used, the magnetic bearings must be able to provide the necessary restraining torques to handle inputs not damped out by shock- and vibration-isolation systems.

In addition to controlling gyroscopic effects, the use of small flywheel modules provides other advantages, Post noted. "It provides very high power densities and reduced gyroscopic moments; it makes it easier to use rotors in pairs and to contain failed rotors. The use of a modular approach also provides redundancy and simplifies manufacturability and transportability," he said.

Given the current state of the technology, "the challenge is not in the components, many of which have been used in other applications," Zorzi said. "The challenge is to match the components to the system and successfully integrate them." By the time the technology is available, however, FES may face tough competition from advanced chemical batteries, fuel cells, super capacitors, and other energy-storage devices currently in development.

"In the end, factors other than technology will determine whether flywheels prove successful," said George Chang, chief of the mechanical energy branch of the DOE Office of Energy Technology in the late 1970s and now a professor of aerospace engineering at Embry Riddle Aeronautical University in Daytona Beach, Fla. "It's really not a matter of technology," he said. "It's an institutional quesion, because it will take strong incentives to switch away from current technology and the industries that have been established to support it. In addition, it will take a major effort to understand the economics of flywheel-powered EVs and to answer questions about the infrastructure needed to make the technology work."