To win a competition, a Georgia Tech student devised a fuel-cell plane to rival today’s best-selling small aircraft
Originally published by Tom Neuman for IEEE Spectrum on May 24h 2016
If you fly one of today’s small piston-powered planes, you will burn many gallons of fuel per hour and suffer through the noise equivalent of a ride on a power mower. But unlike what you face while cutting your grass, if the engine quits, it means an immediate emergency landing at best and a crash at worst. Fortunately, it’s now possible to envision replacing those noisy gas guzzlers of the sky with electric airplanes, which would be considerably quieter, cleaner, safer, and more efficient than today’s aircraft. Indeed, electrification could transform the current small-airplane experience into something vastly more attractive to both pilots and the communities over which they fly.
Recognizing the possibilities—and hoping to spur innovation—NASA recently challenged students to design a four-seat, all-electric aircraft capable of entering service by 2020. After I read an announcement describing the contest, I came up with a design that ultimately won first place in the graduate division of NASA’s competition. My design relies on fuel cells for propulsion and uses an unusual motor placement to maximize efficiency. For me, working on its design was a window into the possibilities and also a vivid lesson about the significant challenges ahead for electric flight.
When NASA issued this challenge in 2014, I was a graduate student in aerospace engineering at the Georgia Institute of Technology, in Atlanta. But this contest wasn’t my first foray into electric flight. In 2008, when I was 17 years old, about six months after the first Tesla Roadsters hit the road, I worked on another electric plane—one with a 9-foot (3-meter) wingspan designed for a remote-controlled aircraft competition.
At the time, electric flight was becoming an increasingly reasonable proposition as lithium-ion batteries became lighter and cheaper. Flying with electric motors made the planes cleaner, quieter, and easier to operate. Still, the technology available in 2008 would have limited pilots to flights of a few minutes. Similar considerations applied to Tesla’s first car: It offered some compelling advantages, but it couldn’t go nearly as far as its gasoline-powered counterparts.
Since then, battery technology has improved to the point where all-electric cars have become an established niche: Nissan introduced the mass-market Leaf at the end of 2010, and Tesla has increased the size of its flagship automobile from the two-seat Roadster to the five-seat Model S. Electric planes have also benefited from these battery advances. In the past two years, both Pipistrel, the Slovenian light-aircraft manufacturer, and Airbus SAS, based in France, have introduced electric two-seat trainers.
While the range of electric aircraft has been growing, limited flight time remains their main weakness. The reason is that batteries are heavy relative to the propulsive energy they provide—by a factor of 10 or more compared with that of gasoline-powered internal combustion. For ground vehicles, designers can compensate somewhat for this shortcoming by adding more or bigger batteries. But aircraft are extremely sensitive to extra weight: Just about every component of a plane’s structure must grow in size for each added kilogram. The requirement for beefier components in turn leads to a heavier aircraft, one that requires still more energy and therefore larger batteries to fly. This vicious circle means that for electric-plane design, adding batteries to boost range isn’t a viable strategy.
Nevertheless, NASA has repeatedly encouraged innovators to try to build electric aircraft that could compete on both range and size with gasoline-powered airplanes. In 2011, the top two aircraft in NASA’s Green Flight Challenge were completely battery-powered and flew for nearly 2 hours at an average speed exceeding 100 miles (161 kilometers) per hour. In an unusual twist, the designers of the second-place winner, the e-Genius, built by researchers at the University of Stuttgart, positioned the electric motor atop the tail. Its distance from the center of the fuselage allowed the plane’s makers to install a large and highly efficient propeller while keeping landing gear short and light.
In 2014, NASA and two industry partners launched a project called LEAPTech (short for leading-edge asynchronous-propeller technology). They’ve constructed a special carbon-fiber wing equipped with 18 electric motors and propellers, which they used for ground testing. They plan to place a similar wing on an Italian-built Tecnam P2006T light aircraft. The multitude of propellers, each potentially adjustable independently, will allow this plane to use a smaller wing than normal, reducing the amount of energy consumed in flight.
In September of that same year, NASA announced a design competition for students that it hoped would push the limits of electric-airplane technology far beyond what previous projects had accomplished. Always looking for the opportunity to tackle a unique design challenge, I signed up without hesitation and ended up as the only one-person team in the contest.
The target NASA had in its sights seemed impossibly distant, and yet it required the designs to be flight-ready in five years (by 2020). NASA’s design objective for the competition was a four-seat airplane with a range of 800 nautical miles (921 statute miles, or 1,482 km) and a cruising speed of 175 knots (equal to 201 mph, or 323 km/h). And while most designers might scoff at such a challenge, seeing these specs as completely normal, the goals were more than seven times as far and twice as fast as the brand new Pipistrel Alpha Electro electric airplane could go. Quick math shows that the NASA performance targets demand about 30 times as much energy as the Pipistrel is capable of providing. What could change so drastically in five years so as to allow that?
Some proponents of electric flight have argued that matching the ranges of gasoline-powered planes is not really necessary. They point out that pilots, like drivers, rarely make use of the full range of their vehicles, so little utility is lost in designing for shorter flights. Think of the Nissan Leaf, which has only one-third the range of a typical car.
That’s all true. But it’s also true that range anxiety is a scarier proposition in a plane than in a car. And certifying and selling a fully electric plane would be enough of a battle, I reasoned, without it having a limited range. So I resolved to meet as many of NASA’s performance objectives as possible. That way, if some manufacturer should decide to build it, the plane could compete head-to-head with existing gasoline-powered models, including the US $500,000 four-seat Cirrus SR22, the best-selling single-engine aircraft of the past decade. That, I knew, was really the plane to beat.
I started my design project by determining whether converting the power plant of an existing gasoline-fueled plane to electric, using either batteries or fuel cells, would work. I knew that batteries lack energy and fuel cells lack power, but I didn’t know which hurdle would be more difficult to overcome.
I started with batteries. They were familiar and flexible—you can go from one cell to 6,831, as the designers of the Roadster did, if needed. I discovered that if I swapped out all 1,020 pounds (462 kilograms) of engine and fuel from a Cirrus SR22 and replaced them with a lithium-ion battery and an electric motor, the plane could fly for about half an hour. That would allow it to travel only about 100 miles (some 160 km). Boosting the range to anything like the SR22’s by adding more batteries proved futile, because of the vicious weight circle I described earlier.
The Vapor Aircraft
This design for a four-seat electric airplane borrows heavily from the fuel-cell-powered Toyota Mirai automobile.
This exercise showed why Pipistrel and Airbus are making trainers and not practical airplanes—there’s not enough energy. Despite this, I gave batteries one more chance, because they should improve in the next few years. The Airbus battery cells provide 200 watt-hours of energy per kilogram. Projections suggest that with substantial technology investment, advanced lithium cells may reach 400 Wh/kg by 2020. Doubling the energy means doubling the range to 200 miles, but that would still be far short of NASA’s 921-mile goal.
As I confronted this reality, I began to think carefully about energy storage. I knew that fuel cells can offer substantially more energy per unit mass than batteries. The trade-off is that to get this energy, fuel cells have low specific power—the number of watts you can get per kilogram. Still, it was reasonable to consider using fuel cells to power an aircraft: In 2008, for example, Boeing converted a Diamond HK36 Super Dimona, a two-seat motor-glider, to electric power using a proton-exchange-membrane fuel cell in addition to batteries.
So I considered the possibility of retrofitting a Cirrus SR22 with fuel cells. The SR22’s normal power plant is a Continental IO-550-N, a six-cylinder, horizontally opposed, air-cooled engine that weighs 187 kg and provides 310 horsepower (231 kilowatts). By removing the engine and fuel and replacing them with a fuel cell of the same weight, I could possibly produce a similar amount of power. But to do that, the fuel cell would have to provide 500 W/kg of specific power. And at that level of specific power, the specific energy of the fuel cell would be about 400 Wh/ kg, roughly as good as the best batteries I could expect to use, which I already knew wouldn’t let my plane fly very far. To provide a specific energy of 800 Wh/kg, the fuel cell’s specific power drops to 200 W/ kg, well below the power needed to fly at 200 mph.
My options seemed to be narrowing the more I investigated the limits of electric energy storage. The only solution was to reduce the energy and power demands of the aircraft. But I knew that the SR22 was already a nearly optimal aircraft made of lightweight composites. Design tweaks wouldn’t cut it.
When an aircraft designer gets stuck between a rock and hard place, there are two ways out, and neither one of them will make the boss happy: Compromise performance, or invest in technology development. I decided to take a closer look at the performance before risking my design on unproven or expensive technologies. But which performance requirement should be sacrificed, I wondered—speed or range?
Some of my earlier analyses had shown that the sensitivity to range was similar to the sensitivity to speed, so my choice of the two had to rely on what it takes to compete in the four-seat general-aviation market. I examined six different aircraft being sold and found that their ranges were all near NASA’s objective of 921 miles, while their cruising speeds differed substantially. This variance led me to aim for the longest possible range and worry less about speed. After a few more calculations, I decided to shoot for 150 knots (173 mph, or 278 km/h) instead of the desired 175 knots.
Vapor and Some Rivals Compared
Reducing the cruise speed to this level diminished energy consumption by 30 percent, and it also slashed the amount of power needed for propulsion. But it still didn’t put my design in any position to meet NASA’s requirements. I hesitated to decrease range too, so I looked for other ways to trim energy consumption that didn’t involve millions of dollars in technology investments.
Fortunately, electric propulsion offers some flexibility that the engineers at Cirrus did not enjoy. Unlike combustion engines, electric motors are compact and efficient. These small, light motors can be placed in many more locations on the aircraft than would be practical for a combustion engine. If applied strategically, this tactic can distribute the power production across more or larger propellers. And the greater the area swept by propellers, the more efficient and quieter they become.
I ran yet another analysis and found a sweet spot in efficiency using two rather large propellers attached to a pair of motors. Instead of mounting them conventionally, on the wing or fuselage, I put them in my design atop the plane’s V-shaped tail, where the airflow is cleaner.
This simple strategy not only improved propulsive efficiency (from 85 to 92 percent), it also benefited the plane’s aerodynamics. Now air could flow more cleanly over both fuselage and wing. And although the propellers were large, putting them on the tail meant that I didn’t have to increase the height (and therefore, weight) of the landing gear. Having short gear made choosing retractable wheels much more palatable, and this reduced drag even further.
When I ran the next analysis, I found that this change, combined with some more optimization, decreased the plane’s energy consumption by another 27 percent. Indeed, this design change had lowered the power demand to the point that it became feasible to fly the plane on hydrogen-powered fuel cells. That’s when I dubbed my V-tailed, hydrogen-powered design “Vapor.”
Once the general parameters of the design became clear in this way, I could work on the details. One was to select the best type and layout for the fuel cells. I examined various fuel-cell systems that have been used in the automotive and aerospace applications and found that the fuel-cell stacks and hydrogen tanks of the type used in the 2015 Toyota Mirai would create a lighter and more compact system than was thought possible eight years ago, when Boeing first flew its fuel-cell powered HK36. What’s more, all of these components would fit nicely into the Vapor’s airframe.
The fuel-cell system I had designed with available technology delivers 800 Wh/kg at 55 percent efficiency. That was certainly better than 400 Wh/kg for the best lithium-based batteries I could expect or with 25 percent efficiency for modern gas engines. Combining fuel-cell power with the plane’s unconventional propeller placement, I was able to arrive at a design that, if mass-produced, could indeed compete with the Cirrus SR22. It would weigh and cost about the same, and its range would be very similar: about 920 miles. The plane’s cruising speed would be somewhat less—173 as opposed to 212 mph. But that seems a reasonable bargain, given that the electric aircraft would consume about a quarter of the energy per flight.
My goal was for Vapor to be attractive to pilots—they are, after all, the ones who buy, fly, and maintain small planes. And I think the design meets that requirement. The decrease in energy consumption and the elimination of the gasoline engine (and all the routine maintenance it requires) will likely reduce operating costs. What’s more, the reduction in noise level, from 92 decibels to 76 dB, should improve cabin comfort considerably. And the very high reliability of electric motors should give both pilots and passengers greater peace of mind.
Given recent advances in fuel cells and electric motors, Vapor, or something like it, could well be built and flown right now. The technology is certainly ripe for exploitation. But it’s unclear how regulatory authorities will react to the advent of such all-electric designs. That’s important because uncertainties in the certification process can doom an effort to develop an aircraft for commercial production.
Another hurdle is that hydrogen has not yet caught on as a fuel for automobiles, much less for airplanes. Both applications suffer the chicken-and-egg problem: Until it becomes a popular fuel there will be little infrastructure to support the distribution of hydrogen, and until there is infrastructure to support its availability, it won’t become popular.
Despite these hurdles, the proposition of an all-electric aircraft flying seven times as far and twice as fast as current designs is exciting. With NASA’s first-place endorsement, it isn’t a stretch to say that the Vapor, or a plane based on its design, could begin taking to the skies by 2020, just as the competition organizers at NASA had intended.
This article appears in the June 2016 print issue as “Fly the Electric Skies.”
About the Author
Tom Neuman is an engineer at Toyota’s Technical Center in Michigan.