EAA - Experimental Aircraft Association  

Infinite Menus, Copyright 2006, OpenCube Inc. All Rights Reserved.



Navigation

Tools:   Bookmark and Share Font Size: default Font Size: medium Font Size: large

EAA Experimenter

[ Home | Subscribe | Issues | Articles | Q&A | How To | Forum Review ]
[ Hints for Homebuilders | Glossary | Polls | Around the Web | Submit an Article]

The Electric-Powered Aircraft

Technical challenges

By David Ullman, EAA 0446096, ullman@robustdecisions.com  

The thought of buzzing around the sky, quietly, with no carbon emissions is an enchanting one – and the promise of electrically powered airplanes. Just plug it in for a couple of hours and go flying. No carb ice, no oil stains, no $5 per gallon avgas. But how close is this to becoming a reality? In this article, the technologies that are needed to design, build and fly electrically powered aircraft are explored. The goal is to describe the different types of components, their state-of-art and what developments we can expect.  In the first article of this series (KITPLANES Oct 2009), I made some predictions about electric aircraft. I predicted that within the next four-and-a-half years, before 2015:

  1. An electrically powered airplane will stay aloft for two hours carrying two people.
  2. An electrically powered airplane will be cost competitive for light-sport aircraft and trainers.
  3. Electric airplanes will be powered by batteries with energy densities at least 50 percent better than those available today.
  4. An electric airplane will land at an airport near you!

I plan to prove to you why I think I’m safe with these predictions.

Every electrically powered airplane will obviously have a motor and batteries. But of equal importance are the controller, the battery management system (BMS), and the airframe matched to the electric airplane needs. These, while not as obvious, are critical to efficient power usage and battery life.

Motors
Let’s start with the motor. To date, most motors being used in electric airplanes are purpose designed. Off-the-shelf motors – those developed for use in industry, electric cars, and model airplanes – aren’t generally a good fit. Typically, model airplane motors are too small (in the .5- to 7-kilowatt (kw) range*), and industrial motors too heavy. Electric car motors are heavy and lack efficiency at peak horsepower needed for an airplane.

*Note that there are 1.34 “horses” per kw. Click here for a kw to hp calculator.
 
To understand what’s being developed for electric airplanes, it’s necessary to understand some motor basics. An electric motor has two major parts, a rotor and a stator. The magnets in the rotor chase the magnets in the stator, causing rotation. The effective location of the magnets in either the rotor or the stator must be moved to keep the chase going.

Part 2 Figure 1
Brushed DC electric motor

The most common type of direct current (DC) motor, a “brushed” DC motor, is shown in the figure above. In the illustration, the magnets in the stator are labeled with their north or south polarity – only four are shown as this is just a representative sample. These magnets can be permanent magnets or created by an energized coil of wire, i.e. electromagnets. Electricity is fed to the rotor through the commutator.

The commutator is designed so that the electromagnets can switch polarity by reversing the current flow. At any time during rotation (with the motor energized), the rotor magnets are attracted to and repelled from the stator magnets in such a way to cause rotation. The commutator has segments all the way around the rotor, and the stator is connected to brushes that rub against (making contact with) these segments to conduct the electricity.

Brushes are a weak point in this design; they cause arcing, they wear, they cause drag, and contaminants like dirt and grease can decrease their efficiency. However, to make a brushed DC motor run, you just need to supply a current to the wires, so no controller is needed unless you want to control the speed. To vary the speed, you need to adjust the voltage accordingly, or as discussed later, the average voltage seen by the motor.
 
Brushless DC motors are electrically commutated, eliminating the problems with brushes. This is accomplished by putting permanent magnets on the rotor and controlling the electromagnets in the stator, changing the polarity to attract and repel them as shown in the illustration below.

t
Brushless DC electric motor

Where this configuration gets rid of the brushes, it adds a controller – both serving to alternate the polarity of the electromagnets and the voltage level. In other words, the current flowing to the electromagnets is alternating direction to change polarity which explains why these motors are often referred to as AC (alternating current) motors. In the past, controllers were of analog design, but all modern controllers use digital signal processing.

Brushless motors typically run cooler and are more efficient than brushed motors, but they cost more. The efficiency of brushless motors has improved, and their size and weight (weight-to-power ratio) have decreased in recent years with the improvement in magnet technology.

Outrunners
Where the motors described so far have had the rotor in the middle with the stator around it, a development unique to aircraft and model aircraft motors are “outrunners”as shown to the right. Outrunners have stationary inner coils and a rotating outer shell. The reasons for doing this are for lightness, to run cooling air through the middle of the motor, and to provide a convenient place to mount the propeller.

The most refined airplane outrunner is on the Antares motor glider. (See image to the right.) Its brushless 42-kw motor is powered by 288 volts at 160 amps. It’s 90 percent efficient.
 
Most motors for electric airplanes are proprietary. Two companies that make off-the-shelf electric motors that may be suitable for electric airplanes are Agni Motors and Perm Motors. The websites for information on these is in the side bar near the bottom of this page.

Brushless DC electric motor
Outrunner brushless DC electric motor

 

Antares motor
The Antares outrunner motor (by permission Lange Aviation).


Batteries and Fuel Cells
Batteries store energy and provide it to the motor via the controller. The most important characteristic of batteries for airplanes is their energy density, usually measured in watt-hours per pound or kilogram (wh/lb or wh/kg), where 1 kg is 2.2 lbs. There are other important factors to consider, but energy density is clearly the most important for airplanes. And in the development of an electric airplane, batteries are the determining factor. The reason we have a recent flurry of activity in electric airplane development is that lithium-ion batteries (sometimes Li-ion, Li, or LIB) were introduced in 2006. They changed everything and continue to do so.
 
A simple example to help in understanding the battery’s importance: Say you want to electrify your Pietenpol. You want a 65-hp engine equivalent (49 kw), and you want the airplane to weigh the same as if it had the original engine, fuel tank, and full load of fuel. If the electric motor and controller weigh about 50 lbs (22.7 kg), this leaves about 300 lbs (136 kg) for batteries. If you use lead-acid batteries with an energy density of 30 wh/kg (see the table below), then your batteries could supply 4,000 wh or 4 kwh. So, with your 30-kw motor, you can get about 8 minutes of power – just beyond the end of the runway.

Before you think too much about this example, remember that the electric motors weigh just a fraction of a typical internal combustion engine, the batteries replace the fuel, and the Pietenpol is probably not the ideal electric airplane. Additionally, and this is critical, the energy density of batteries is key to the development of a viable electric airplane, and lead-acid batteries aren’t very good for this application. We’ll see that with lithium-ion batteries things get much better, over an hour flight, as will be discussed later.

The table below lists major battery types, their energy densities, and their costs. What is important here is that newer battery types like lithium-ion (Li) have energy densities three to five times greater than good-old lead acid batteries and are improving rapidly. There are many different chemistries used in Li batteries, hence a wide range of values. Also shown in the table are the batteries used in the Tesla Roadster (a $130,000, 0 to 60 in 3.9 seconds sports car), a lithium formulation that’s a good benchmark because it’s from a production vehicle.

Cost chart
Cost chart

As impressive as the numbers are for lithium batteries, their energy densities are increasing at 8 to 10 percent per year, a doubling every 10 years. The limit on this growth is the theoretical maximum of 5200 wh/kg. This clearly will never be achieved, but 365 wh/kg is expected in the next couple of years, a full ten times better than lead acid. Even better, other formulations are being developed, and so this density will surely be improved in the future.
 
The price of Li batteries has been very high but is falling rapidly. I recently helped build an electric car with $15,000 in batteries in it. Note that the high-end price on the table includes the battery management system, which is discussed in the next section on controllers. Currently most Li batteries come from China. However, General Motors announced last August that it’s investing $43 million for a battery manufacturing plant in Michigan.
 
On the last line in the table, the energy density of aviation fuel is included as a reality check. Gasoline is very energy dense, but before you write off batteries as being too heavy to ever be a realistic energy source for an airplane, consider that electric motors are 90 percent efficient, and internal combustion engines are only about 20 percent. The last column of the table shows the energy densities corrected for these efficiencies. Clearly, batteries have a long way to go to rival avgas on energy density alone, even at 20 percent efficiency.
 
Batteries aren’t used alone but are wired together to get high voltage. Higher voltage gives higher efficiency. (This is the same reason power companies transfer power at very high voltage.) Also, you can charge faster at high voltage because you’re at a lower current. To give you some feel for this, consider this example from the electric car I recently worked on. The battery pack consisted of 100 lithium-ion batteries. Each battery is 3.2 volts, so the total is 320 volts. These batteries weigh 7 lbs each for a total weight of 700 lbs (320 kg). They have an energy density of 140 wh/kg, so there’s a total of 45 kwh. This battery pack in an electric car is equivalent to about 1.2 gallons of gas at 100 percent efficiency or 6 gallons realistically. It’s also worth noting that the batteries used in this example cost $150 each, a total cost of $15,000.

Other Energy Options
Two additional sources of electrical energy that need to be mentioned are solar and fuel cells. It may be practical to use solar cells to supplement the energy in an electric airplane. Current solar collectors generate about 11 watts per square foot (sqft). For a 150-sqft wing area, this is about 1.6 kw or a little over 2 hp continuous. For a very clean airplane like the Yuneec (see below), this can extend the flight time by about 15 percent, which is why Yuneec is planning on solar cell additions to their Model e430. Additionally, the efficiency of solar cells is increasing, but current efficiencies are only a fraction of their ideal efficiency of 100 watts per square foot. At least there’s plenty of sunshine above the clouds.

r
Battery pack in an electric car showing 64 of the 100 batteries in the pack.

Fuel cells are another approach for obtaining electrical energy. They’re effectively mechanically rechargeable batteries. Batteries have all their chemistry inside, whereas chemicals constantly flow into fuel cells. Most fuel cells today use hydrogen and oxygen as the chemicals. The efficiency of fuel cells is about 20 to 30 percent when converted to useful work. As they develop, they might compete with lithium-ion batteries. The SkySpark project in Italy is combining fuel cells with batteries for a hybrid aircraft propulsion system to test this opton. Of course Boeing has already proven that a hydrogen fuel cell with a Li battery pack can power a full-size aircraft (the Diamond Katana) back in 2007.

Regardless where the electric energy comes from – batteries, fuel cells, or solar – a controller is needed. As with batteries and motors, controllers have changed a lot in the recent past.

Controllers
When I first started learning about electric-powered vehicles, I had no idea what a controller did and how it worked. My experience centered on using a rheostat, a big variable resistor to control DC motors. Of course rheostats are hopelessly inefficient (key on the word “resistor” in “variable resistor”), and what is used is much more sophisticated.

Controllers not only change the battery energy flow in response to the throttle setting, but they protect the motor and batteries from any spikes, shorts, or other anomalies that might occur in the system. Basically, all controllers use pulse width modulation to control the energy to the motor. The controller turns off and on the voltage supplied by the batteries as pulses, typically at 15 kilohertz (kHz) or 15,000 times per second. Three different sets of pulses are shown in the above graphic. The inductance in the system acts like a damper so the motor actually sees a very smoothed-out current supply with a trivial 15-kHz ripple. In other words, the average voltage seen by the motor at, say, a 10 percent duty cycle on a 100-volt supply is a slightly wavy 10 volts. The waviness is so small in proportion to the average voltage and the inertia in the system is so high, the motor runs a virtually constant speed.

f
Pulse width modulation

At this time the only commercially available controllers are those developed for cars or motorcycles. The best of the lot is the Zilla, manufactured by Cafe Electric, currently restructuring. The Zilla and others are designed for high-peak loads as when the car is accelerating and thus are not optimized for airplane use. Most aircraft to date have custom controllers, but this is expected to change as the field matures.

Battery Management Systems
Lithium-ion batteries require a battery management system (BMS) to control the energy flow to and from the individual batteries, keeping them balanced while monitoring their temperature. To explain why this is needed, consider a simple system of two batteries in series. If wired to a motor, current will flow through both batteries as they give up energy. But the batteries won’t be identically the same, so one battery may give up energy faster than the other. Thus, when the battery that’s giving up energy the fastest is empty, the motor will stop even though there’s still energy left in the other one. It’s similar to when one battery goes dead in your flashlight.

The same can happen when the batteries are being charged. One can charge faster than the other and may overcharge (damaging the battery) while the other is still not up to its potential. Therefore, the job of the BMS is to monitor the state of each battery and ensure that all are being treated equally to optimize the system.
 
As with controllers, BMSs are still under development, each system unique to one another.

Airframes
The last component to consider is the airframe. To date, most electric airplanes have been converted motor gliders with purpose-built airplanes just beginning to appear. (The Antares 20E motor glider is the exception having been on the market at $300,000-plus for about five years.) As should be obvious by now, energy is limited, and so an efficient airframe is necessary. To make this article easier to relate to current aircraft, I developed a simple model from three different aircraft (details in the sidebar at the end of this article).

The first is a J-3 Cub. It isn’t an ideal candidate for an electric airplane, but it will make for a good starting point. For this airplane, I assume that:

  • the engine, fuel tank, and all other power components were removed
  • an electric motor, controller, and BMS were installed and they weigh 50 lbs
  • the airplane carries a pilot and a passenger
  • 300 lbs of lithium-ion batteries producing 140 wh/lb are installed to bring the weight back to the same as the original with full fuel tanks (1220 lbs gross)
  • the airplane wastes no energy on the ground (starting point on the runway), takes off and climbs to 2500 feet, and then cruises at that altitude until it runs out of energy.

The Li batteries for this airplane store 19.1 kwh. If the Cub is flown at its recommended cruise speed of 65 mph (56 knots), it will be able to fly for 0.76 hour, which is about 45 minutes. If you throttle back to just overstall, then you can get over 1 hour of duration. To make this conversion would cost $10,000 to $15,000 for the batteries alone. But you would never have to purchase avgas again.
 
The second is the Yuneec e430, a purpose-built electric airplane that was recently introduced and is still undergoing test flights. It’s similar in size to the J-3 Cub, holds two persons, and has a gross weight of 1034 lbs. The Yuneec e430 is composite and very clean aerodynamically, and it comes with either a 6-pack of batteries or a 10-pack. The larger pack weighs 286 lbs (130 kg) and is estimated to store 18 kwh. It has a recommended cruise of 60 mph (52 knots). At this speed, my simple model shows it will fly for 2 hours. Company literature says that it estimates 2.25 to 2.5, but the Yuneec hasn’t flown that long as of this writing. Sales are expected this year.

The Yuneec e430 (by permission Yuneec International)
The Yuneec e430 (by permission Yuneec International)

Why is the Yuneec so much better than the Cub? It has much lower drag so the power is used more efficiently. As shown in the sidebar, drag is composed of induced drag and parasitic drag. The Yuneec is designed somewhere between a Cub and a motor glider – a plane designed to have both drag components as low as possible. Induced drag is directly proportional to aspect ratio (the wingspan divided by the chord). The Cub’s is 7:1 and the Yuneec’s is twice that at nearly 14:1.

The electric airplane with the most sales to date (about 50) is the Antares 20E motor glider and is more akin to the Yuneec than the Cub. These have enough stored energy to climb from sea level to about 10,000 feet although typical operation is to climb to the safest altitude for the flight and go thermal or wave hunting, adding power only when the altitude gets too low between lift sources. The Antares is a true self-launch sailplane with an aspect ratio of 31.7:1. Both the Yuneec and the Antares have very smooth skins and are very clean, minimizing parasitic drag, the polar opposite of the Cub.
 
Since flight time is directly proportional to the energy stored, and the energy stored is proportional to the weight of the batteries, another metric to consider is that portion of the gross weight used for batteries. For the Cub, 29 percent of gross is batteries, whereas the Yuneec (with its larger battery pack) uses 33 percent. This begs the question, is there an airframe that can handle a higher percentage of battery weight?
 
One potential approach is to see what happens with a larger airplane, like an RV-10. If you replace the engine and fuel system with 80 lbs of electric motor and controller and 1120 lbs of batteries, keeping the gross the same with two passengers, you could potentially have an “RV-10e.”

So the third example in this study, the RV-10e would have 44 percent of its gross taken up by batteries. This plane can store 71 kwh of energy, nearly four times the Yuneec or the Cub, but it takes more power to fly (hence the higher assumed motor weight). My simulation shows that it does quite good if you fly it slow. At 70 mph, it will stay up for 1.8 hours. If you crank it up to its cruise speed of 170 mph, then the duration falls to 0.6 hours. (However, if the 10e was streamlined and given a high aspect ratio wing, then…)
 
These results imply that going bigger may not be a bad approach for an electric airplane. One drawback, however, is that the batteries for the RV-10e will cost about $40,000 to $60,000 at today's prices, which is marginally competitive with the price of a new IO-540. It may be that as the price of batteries comes down and their efficiency increases, a larger airplane begins to makes sense. This got a friend and me to think about converting a DC-3 and using large AC motors – then we would have the “AC-DC-3.”
 
Conclusions
The electric airplane industry is clearly in its infancy, but developments are happening fast. Its future depends on the cost and energy density of batteries. Even now, with the current cost of avgas nearing $5 per gallon, the Yuneec will fly virtually for free after a couple hundred hours if you don't factor the life of the battery. As batteries get better and cheaper and purpose-built planes are designed, I have no fear that my predictions won’t be met. Listen for the hum near you.

Sidebar
To calculate the performance for the three aircraft in a uniform manner, I computed the maximum amount of stored energy for each by first subtracting the weight of the engine, fuel, and payload from the advertised gross weight and adding in the estimated weight of a motor and controller (50 lbs for the smaller aircraft and 80 lbs for the RV-10e, both educated guesses). The remaining weight is assumed batteries with energy density of 140 wh/kg.
 
To compute energy use for constant level flight at different speeds, I estimated the power versus velocity curve for each aircraft. I assumed a general equation for power used to overcome parasitic and induced drag. Also, I assumed that cruise velocity data published for the airplanes was at 65 percent power and flight at a speed just above stall was at 30 percent power. I added basic parasitic plus the induced drag equation to these two points with the following results.

Where I had other information, I verified these equations as best as possible. They’re certainly accurate enough for the comparisons made in this article. Note that power (P) is horsepower and velocity (V) is miles per hour.
J-3 Cub P = .0000835 * V3 + 426 / V
Yuneec P = .0000505 * V3 + 36 / V
RV-10e P = .0000201 * V3 + 2880 / V
When computing the time aloft, keep in mind that the energy stored in the batteries is used to lift the weight of the airplane to 2500 feet, and then the planes use power based on the equations until the energy is used up, which is the time of flight reported in the article.
 
Sources
The Antares – www.Lange-Aviation.com/htm/english/products/antares_20e/antares_20E.html
Agni Motors – www.AgniMotors.com
Perm Motor– www.Perm-Motor.de/index.php?id=4&L=1
Yuneec – http://YuneecCoUK.site.securepod.com/Aircraft_specification.html
Zilla – www.CafeElectric.com
Thanks to Otmar Ebenhoech of Cafe Electric and Dean Sigler for their help with the details of this article.

David Ullmanis a retired mechanical engineering professor and author of books on mechanical design and decision making. Hes building a Velocity SE-FG and expects to be flying in 2011. He previously expected to be flying in 2009.

David’s Velocity in its hangar
Current state of David’s Velocity in its hangar, or rather garage, which is 15 feet off the ground; he lives on a steep hill. Never hurts to get a running start.

 
Copyright © 2014 EAA Advertise With EAA :: About EAA :: History :: Job Openings :: Annual Report :: Contact Us :: Disclaimer/Privacy :: Site Map