I remember distinctly the cool December day in 2013 at my company’s headquarters in the Santa Cruz mountains when we met with researchers from NASA to plan tests for a novel propeller configuration for electric aircraft. Somehow, our Joby Aviation team, the NASA researchers, and colleagues from another small California business brainstormed our way into a much more ambitious program than any of us had expected when the meeting began.
Instead of building and testing a scale model, we decided to construct a full-scale wing—one large enough to lift a four-passenger aircraft. And it would have not two or four, but a dozen or more separate motors and propellers arrayed along the leading edge. We could have tested a smaller wing by mounting it on a pickup truck. A full-scale wing would require something a lot more elaborate. And the project would need to be completed in less than a year, on a budget small enough to make most companies turn tail. But we were committed.
Our novel configuration was based on an old concept: The idea—known as a “blown wing”—was to propel air at high speed over the wing using many motors and propellers mounted along the leading edge. Usually, the speed of this airflow is about the same as the speed the aircraft is moving; that’s why airplanes need to pick up speed before they can take off. But with many propellers blowing air over it at high speed, the wing behaves as though it’s traveling faster than it actually is, providing greater lift.
That’s a key advantage, because with greater lift, you can use a smaller wing, one that would otherwise require inordinately long runways so that the plane could take off and land at high speed. The situation is different in flight, when the plane is traveling fast and only a small wing is needed to provide the requisite lift. During that phase of flight, a larger wing is a disadvantage, because drag forces act over the whole area of the wing, reducing efficiency.
So what’s an airplane designer to choose: a big wing or a small one? Takeoff and landing considerations usually rule the day, so aircraft end up with wings that are too large for efficient cruising. A larger wing also means that the aircraft will be tossed around more when it encounters turbulence.
The blown wing provides a solution to this conundrum. During takeoff and landing, air can be blown over the wing at higher speeds, providing additional lift without sacrificing cruise performance. Although a few aircraft have been developed in the past with blown wings, the use of combustion engines for propulsion limited how far their designers could go. They had to use relatively few, large propellers, which aren’t well suited to pushing air at high speed.
It would be more effective to distribute a large number of small propellers across the span of the wing, but for most of aviation history, that arrangement has been impractical.
The problem is that the efficiency and specific power (the ratio of power output to weight) of combustion engines plummet as they’re scaled down. So using a large number of smaller engines results in a less efficient and heavier aircraft.
What’s more, combustion engines are complex beasts. So placing a large number of them on the wing would create a maintenance nightmare. True, a series of propellers could instead be driven by a system of driveshafts and gearboxes connected to a single engine or perhaps a small number of them. But that approach, too, would create additional maintenance concerns and force various design compromises, as the French firm Breguet Aviation discovered in the 1960s with its short-lived 941 model, which used a blown wing.
What has changed the picture, of course, are recent advances in electric propulsion. Electric motors don’t give up much efficiency or specific power as they’re scaled down. And they’re extremely simple—often having just one moving part—so they require very little maintenance. As a result, there is little disadvantage to using a large number of small electric motors, which can be placed at locations on the aircraft where a combustion engine would be impractically bulky or heavy, such as near the wingtips.
Although electric motors can be driven by a combustion-powered generator, the benefits are even greater if the aircraft is battery powered. Indeed, battery-electric propulsion is about three times as efficient as a typical combustion-engine power train. It’s also a lot quieter. And because electricity costs much less than aviation fuel, this two-pronged attack—adding a more efficient power train, plus a more efficient airframe due to the smaller wing—promises to slash operating costs, especially considering the reduced need for maintenance.
So why aren’t all aircraft battery powered? Because, of course, batteries aren’t yet up to the task. Even today’s best are very heavy for their energy content, which severely limits the range of electric aircraft. And they are sometimes prone to catching fire, which some commentators speculate may have been the cause of a fatal crash of an electric airplane in Hungary this past May. But battery technology will no doubt improve with time. So NASA, Joby Aviation, and many other companies are busy exploring various strategies for designing electric aircraft. And reviving the blown wing is one of them.
Five years ago, engineers at NASA started to think about using a large number of electric motors to create a blown wing, later naming the project LEAPTech, for Leading Edge Asynchronous Propeller Technology. (The “Asynchronous” part of that moniker refers to the possibility that the propellers would not necessarily all be spinning at the same speed.)
Joby Aviation, a startup formed in 2009 to develop personal electric aircraft, had already been collaborating with NASA. When my Joby colleagues and I learned about LEAPTech, we jumped at a chance to get involved. Rounding out the LEAPTech collaboration was Empirical Systems Aerospace (ESAero), another small business that had worked with NASA to investigate how electric propulsion can improve aircraft performance.
NASA hoped to vet the idea with an actual test of a wing and propellers, in part because the relevant aerodynamic effects are very complex, and so computational fluid dynamics, or CFD, simulations of them would perhaps not be completely trustworthy. Another concern was that this distributed propulsion system might turn out to be too complicated to operate reliably in a real-world environment.
The test NASA was envisioning would show whether a smaller-than-normal wing with electrically powered leading-edge propellers could produce enough lift to allow a four-passenger airplane to take off at a reasonable speed. Typically, such a test would be done in a wind tunnel. But leasing such a wind tunnel would have exceeded NASA’s small budget for the project. Besides, the waiting lists for appropriately sized wind tunnels were just too long.
So we decided that we would test a prototype LEAPTech wing by mounting it on a truck and driving at a high enough speed to analyze takeoff and landing performance. Such a test is not without precedent. Perhaps most famously, Scaled Composites performed a similar test of the tail of its SpaceShipOne space plane, a method its engineers jokingly also dubbed CFD—for Creative Ford Driving. And Joby had been conducting similar testing for years, with a Ford F-150 Lightning pickup.
Shortly after the fateful 2013 meeting when we decided to build and test a full-scale LEAPTech wing, we divvied up the labor. Joby would work with NASA on the design, also building the wing, motors, and propellers, and it would modify a suitable truck for testing. ESAero would do the wiring, configure the needed instrumentation, and troubleshoot the test setup.
NASA’s initial design sketches featured a wing with 10 leading-edge propellers for takeoff and landing, plus two separate propellers mounted on each wingtip to power the aircraft after takeoff. Putting propellers on the wingtips—where they can reduce drag by counteracting the wingtip vortices—is another old idea that would rarely be practical without electric propulsion. Combustion engines are just too large and heavy to build into a wingtip, and using driveshafts and gearboxes in a wing to turn propellers at the tips creates engineering headaches, just as it does for leading-edge propellers.
After a few months analyzing the problem, we arrived at a design for a wing with a span of about 9 meters and an area of about 5 square meters. It would have a series of 18 propellers, each about a half meter in diameter, distributed along the length of the wing. Altogether, the 18 motors offered some 225 kilowatts, or 300 horsepower.
Although this wing would be used only for ground tests, we designed it with a particular application in mind: an experimental aircraft based on the Tecnam P2006T four-seat twin-engine propeller aircraft. We chose the P2006T because it was a good size, because it had wing-mounted engines (meaning replacing them with electric motors would be straightforward), and because the management at Tecnam was excited about the project.
The experimental aircraft we envisioned would weigh about 1,400 kilograms and take off at a speed of 61 knots (113 kilometers per hour) while cruising at 174 knots (322 km/h). Only the wingtip propellers would be used after the aircraft was up and away. And the leading-edge propellers would be needed just during takeoff and landing. We therefore designed the latter so that their blades could fold flush against their nacelles during the remainder of the flight, making them similar to the folding propellers used in some modern motor gliders. But because our testing would be limited to measuring takeoff and landing performance, the test wing would include neither the wingtip propellers nor the folding mechanism.
These specifications make our design comparable to that of four-seat propeller planes, but with a much smaller wing. Indeed, our wing would be only about a third the size of those on conventional aircraft. On paper, anyway, it would still provide enough lift for normal-speed landings and takeoffs. Our charge was to prove that this surmise matched reality.
For that, we purchased a Peterbilt truck—the kind of thing you might see barreling down the highway with a trailer in tow. On it we constructed supports to mount the wing high enough to minimize the aerodynamic effects of the ground below. To reduce vibration, we attached the wing to the truck using four beefy airbags. The giant winged truck looked distinctly odd, but it was exactly what this job required.
After the design and construction work was complete, we began our tests on the dry lake bed at NASA’s premier flight-testing facility, the Neil A. Armstrong Flight Research Center, at Edwards Air Force Base in California’s Mojave Desert. Tom Wolfe’s 1974 book The Right Stuff and the 1983 movie of the same name made this locale famous. It’s where Chuck Yeager first broke the sound barrier in 1947, and it was the original landing site for the space shuttle.
We were using carefully groomed sections of the lake bed that are maintained as backup runways for the flight test programs currently under way at Edwards. Although it would never leave the ground, we had to treat our unconventional test platform like an aircraft and take all the same precautions to minimize the chance that we’d harm the lake bed or leave behind debris that could later damage an aircraft making an emergency landing there.
Once we had all the batteries and power cables secured and our instrumentation system logging data, we began our tests, which entailed driving the truck at speeds up to about 130 km/h (80 mph) with the wing canted at different angles and with the propellers set to spin at various speeds. A wind tunnel would have offered carefully controlled conditions, whereas we had to estimate our airspeed based on the ground speed of the test vehicle and the wind speed as measured by several weather stations we had placed around the dry lake. To minimize errors and variations, we began at daybreak when the winds were calmest. We also had to find days when other aircraft were not likely to need our runway for an emergency landing, which meant a lot of waiting while NASA tested its X-56A drone and the Air Force tested the Lockheed Martin F-35 Lightning II fighter.
After two months in the desert, we had collected enough data to fully check our computer simulations. We were happy to see the expected performance boost. Indeed, the tests indicated that our predictions for the lift force that could be generated were somewhat conservative. Our electrically blown wing indeed worked!
Based on those encouraging results, NASA decided to further explore the blown-wing concept with a new experimental aircraft, one based on the same aircraft we’d investigated during the LEAPTech project, the Tecnam P2006T. It would be dubbed the X-57 Maxwell, the first piloted NASA X-Plane in more than a decade.
For the X-57, we modified the design in various ways. For one, the X-57 will be using a slightly larger wing. That change would provide enough interior volume for installing the wiring. But a more significant motivation was to improve “loiter” performance: Although the energy required to travel a given distance increases with a larger wing, the energy required to stay in the air for a given amount of time actually goes down. This is important when, for example, the aircraft must circle an airport while waiting for the weather to improve to land.
We also decided to reduce the number of leading-edge propellers from 18 to 12, which we felt would be a better compromise between simplicity and performance. Also, the takeoff speed was decreased slightly to 58 knots (107 km/h), which is more like that of comparable aircraft. And the two wingtip propellers, which we had designed for a “pusher” configuration, were moved from behind the wing to ahead of it, to provide additional ground clearance on landing, when the nose of the plane comes up.
Construction on the X-57 Maxwell is now under way. The original Tecnam P2006T will be modified in stages. For its first flight, probably less than a year away, the two wing-mounted engines will be replaced with two electric motors, without otherwise modifying the wing. The next phase will swap out the original wing for a much smaller one, with the two electric motors moved outboard to the wingtips for greater efficiency. (After this modification, the plane will require longer runways to take off and land.) The final phase will add 12 smaller electric motors spaced along the leading edge, to allow it to take off and land on typical runways while retaining the efficiency gained with the smaller wing.
Flight tests of the X-57 will help NASA engineers gauge the performance and practicality of this configuration. Those tests will also help guide designs for the next generation of distributed electric propulsion, which is soon to arrive. My Joby colleagues and I have already completed a study that examines the possibility of applying similar principles to an 11-seat airliner [PDF].
Wingtip propellers and blown wings are not the only strategies newly made practical by advances in electric propulsion. As another example, my colleagues at Joby and I are developing a five-seat electric aircraft that uses tilting propellers to take off vertically and then transition to normal airplane flight, allowing it to cruise much faster and more efficiently than a helicopter.
Most of today’s airplanes and helicopters look very similar to models from many decades ago, but as this work demonstrates, that’s about to change. Thanks to the flexibility of electric propulsion, aviation is about to experience the greatest renaissance in design since the advent of the jet engine. So be prepared, and don’t forget to fasten your seat belt.
This article appears in the August 2018 print issue as “Reinventing the Wing.”