Autorotation for Electric Rotorcraft: The Recipe

I’ve written a couple of articles about the importance of adapting autorotation technology to the modern era of electric rotorcraft. In September 2019, I talked about the safety benefits of autorotation for eVTOLs, and in January 2023, I talked about safety benefits of autorotation for small, uncrewed aircraft. The quick summary is this: Helicopters have supported autorotation as a basic safety feature from day one, and electric rotorcraft can also support autorotation, to protect people in the air, and to protect people and property on the ground. Today, the US Patent Office has issued our first patent for autorotation technology, and I’ll discuss the basic recipe to achieve controlled autorotative descent for an electric rotorcraft.

Aerodynamic Requirements

Classic helicopters have swashplates for cyclic control. Modern electric rotorcraft usually eliminate the swashplate, though this doesn’t mean they cannot support autorotation. The basic aerodynamic requirements for autorotation are low disk loading (less than 10 lbs/ft²) and adjustable collective. Many existing eVTOLs satisfy these requirements, though few multi-rotor drones do. Low disk loading improves the efficiency of a rotorcraft in hover, while trading off the efficiency in high-speed forward flight. While there seems to be an industry trend toward flying as fast as possible with electric aircraft, I’m a bit of a contrarian. For many applications, given the energy density of today’s lithium-ion batteries, you’re better off slowing down, flying quieter, flying farther, and staying in the air longer. Low disk loading is the key. Adjustable collective has performance benefits as well, though it certainly comes with added cost and weight.

First Phase: Entry

Autorotation can be applied to electric rotorcraft having any number of rotors, though the discussion today will be about an exemplary crewed quad-copter. As a “bad day event,” let’s imagine that the primary battery system suddenly fails, yet the onboard flight controller remains operational, powered from a small secondary battery system. The first thing that needs to happen is dropping collective pitch. Depending upon the inertia within the rotor systems, rotor speed will degrade to a point of being irrecoverable within a matter of seconds. As the math of rotor inertia and shaft power scales, the time available to enter autorotation shrinks in proportion to the rotor diameter, requiring an automated means to decide when to drop collective for an electric rotorcraft with small rotors. In a traditional helicopter, the human pilot hears the engine-out horn, observes the slowing of the main rotor, and instinctively drops the collective, entering autorotation. For the electric quad-copter, the flight controller will detect a catastrophic loss of power, and immediately drop collective on all four rotors. With the proper choice of blade incidence, the rotors will keep spinning at a rate similar to powered flight. To an observer on the ground, it may not be obvious what has just happened. The airflow during powered flight in from the top of the rotors, maintaining altitude by propelling a bunch of air. Once autorotation is initiated, the airflow is coming from the bottom of the rotors. The rotors continue spinning by extracting a portion of the gravitational potential energy (m*g*h) being consumed upon descent.

Second Phase: Glide

Once the quad-copter is in autorotation, the first piece of business is adjusting the forward airspeed to minimize the rate of descent. It’s good to have plenty of time to find your landing spot on the ground. As with a helicopter, that means flying at the airspeed associated with the best rate of climb in powered flight. With all four rotors providing a similar parachute-like effect, we need to either pitch forward to speed up, or pitch backward to slow down. This can be accomplished in one of two ways. A differential in lift between the front and rear rotors can be achieved by either increasing collective of the rotors, or by electrical braking of the motors. Increasing collective momentarily increases the lift of a particular rotor. Increasing electrical braking decreases the lift of a particular rotor. These control methods emulate a swashplate that provides cyclic control in a classic helicopter.

Now we need to find a good landing spot, and hopefully we can approach it into the wind. Adjusting our airspeed required pitching the quad-copter forward or backward. To navigate toward the preferred landing location, we likely need to make both roll and yaw adjustments. The control of roll is like pitching. A differential in lift between the left and right rotors can be achieved by either increasing collective or by electrical braking of the motors. Yaw control is the tricky part. If the four rotors of the quad-copter are coplanar, adjustment of the collective does not induce a required moment to control yaw. In this case, the only way to effect yaw control is through electrical braking, as this induces a counter-torque on the fuselage equal to the braking torque on the rotor. Similar to yaw control of a powered quad-copter, the relative braking of the clockwise and counter-clockwise rotor pairs allows a dynamic adjustment of the net torque acting upon the electric rotorcraft. The flight controller can thus achieve yaw control of the quad-copter during autorotative descent by either braking the clockwise rotors to turn right, or the counter-clockwise rotors to turn left. If the four rotors of a quad-copter have a few degrees of dihedral along the fore-aft axis, then collective adjustment may be sufficient to provide yaw control, as the tilted angle of the rotors allows a component of the thrust vector to induce a yaw moment relative to the center of gravity.

It’s interesting to consider the autorotative regions for the four rotors during controlled descent. As with a normal helicopter, each rotor achieves an equilibrium rate wherein the power extracted by the driving region of the rotor (shown in green) balances with the power consumed by the driven (shown in blue) and stalled regions (shown in red). If the quad-copter were in a vertical descent, the three regions would be centered on the rotor hub. However, in forward flight, which minimizes the rate of autorotative descent, we observe the regions offset toward the retreating side of the rotor disk, as we do for a normal helicopter. This reflects the dissymmetry of lift resulting from the air over the advancing blade being faster than the air over the retreating blade. When electrical braking is applied to a particular rotor, we have added another element to the power equation. To maintain power equilibrium, the rotor rate will slow, the stalled region grows, and yet the driving region grows enough to extract power to balance with the driven region, stalled region, and electrical braking. If too much braking is applied, an equilibrium is not maintained, and the rotors fail to remain in a windmilling state. Similarly, when collective is increased for a particular rotor, we observe a slowing of the rotor rate. However, in this instance, the stalled region shrinks with increased blade incidence, while the driven region grows. Too much collective pitch can also break the autorotative balance of power.

Third Phase: Flare

This is the fun part for a pilot. (When an autonomous flight controller does the work, it’s less fun.) As the quad-copter approaches the ground, it’s desirable to arrest the rate of autorotative descent as well as the forward airspeed. For our crewed quad-copter with low disk-loading, a typical rate of vertical descent is 19 mph (1,700 ft/min), while having a forward airspeed of 70 mph. With careful timing relative to the approaching terrain, the quad-copter pitches backward, using either differential collective or differential electrical braking. The angle of the “flare” is chosen to simultaneously arrest both the forward airspeed and vertical descent rate toward zero velocity. A large amount of energy is suddenly absorbed within the rotor systems, causing the rotors to markedly increase in speed. The timing of when to begin the flare is critical, as the goal is to achieve zero velocity, or something close to it, when our exemplary quad-copter is within 20 to 30 feet of the ground. Once the rate of descent is fully arrested, the rotor planes are brought into a level state.

Fourth Phase: Touch Down

I’ve been talking about electrical braking in the last two phases, though I haven’t addressed the question of where to put the extra energy. For an electric rotorcraft having a sizable secondary storage battery, or perhaps a situation wherein the primary storage battery was simply drained, regenerative braking from the entry into autorotation can yield a valuable recovery of energy. Once the quad-copter has reached a stationary position above the terrain, it’s feasible to initiate a power recovery, allowing the rotorcraft to hover for an extended period, perhaps taxing to a more desirable location before eventually landing. In other quad-copter designs, however, electrical braking results in the extra energy being dumped and dissipated into a resistive load. In these cases, the touch down phase of autorotation involves the unloading of the inertial energy within the four rotors. Collective is gradually increased across all four rotors, with some degree of differential adjustment for attitude control, and the quad-copter comes to a cushioned landing on the ground as the rotors gradually lose speed.

More to Come

My colleague, Jeff Bernstein, and I received US Patent 11,634,235, “Electrically Powered Rotorcraft Capable of Autorotative Landing,” which describes the methods I’ve outlined in this article. We have a follow-on patent application that describes a simplified design of an electric rotorcraft to support autorotation, without the need for electrically-actuated collective controls. I’ll share these methods in a future article. If I’ve convinced you of the value and feasibility of autorotation for electric rotorcraft, you’ll enjoy the simplified design.