AeroVelo, Inc.
Aerovelo's Atlas, the human-powered helicopter's quad-rotor design.
Aerovelo's Atlas, the human-powered helicopter's quad-rotor design.
Aerovelo's Atlas, the human-powered helicopter's quad-rotor design.
Aerovelo's Atlas, the human-powered helicopter's quad-rotor design.
Aerovelo's Atlas, the human-powered helicopter's quad-rotor design.

Inside the Sikorsky Prize-Winning Human-Powered Helicopter

Aug. 28, 2013
An engineer behind a human-powered helicopter describes its construction.
The Igor Sikorsky Human-Powered Helicopter Award (a $250,000 engineering prize) was unclaimed for 33 years, but the University of Toronto recently claimed the prize with an aircraft called Atlas, which weighs 121.4 lbs., and spans 162 ft. The record-setting, pedal powered flight lasted 69.3 seconds and reached an altitude of 3.3 m (10.83ft). The U of T engineering team was led by former students Cameron Robertson and Todd Reichert, who piloted the Atlas. But U of T was an underdog. The University of Maryland was favored to win because it has what many consider to be the best rotorcraft school in the U.S. U of T built and tested the Atlas in just eight months. The majority of the parts for Atlas had to be custom made, so the Canadian students created a separate company called AeroVelo, Inc. to handle part manufacturing. The aircraft made its historic flight in Ontario inside a soccer arena. Here, Robertson speaks about the challenges and success of Atlas.

Machine Design: What drove your team to compete for the Sikorsky prize?

Cameron Robertson: Our impetus to get involved was Gamera, the helicopter made by the University of Maryland team. They had been working on this project since, I believe, early 2009. So for a long time in this race for the prize, they were viewed as the team that was closest, and they were a pretty large team. It seemed like a David and Goliath competition.

What were your other challenges?
We effectively had zero helicopter background. To a lot of outsiders it seemed as though we didn’t have what it would take. It was pretty interesting starting from square one.

How were you able to build a prize-winner?
We used a technique called multidisciplinary optimization. This is a computational approach that looks at the aerodynamics and structures and helps come up with the best overall solution. That includes the aerodynamic and structural analysis. We program the aerodynamic code, we validate it, and we have confidence in it. Anything we design using that code will be fairly accurate. So it allows us to skip the actual testing phase and go directly from design to prototyping.

Describe the structure of the rotors.
Our rotors spin incredibly slowly at about 10 RPM. Our rotors are 10 m. long. We picked that size because it was the most we could comfortably build. It minimized the technological unknowns, but in fact, the perfect human-powered helicopter of our configuration would be another 30% bigger. The good news is that the extra 30% only gets you maybe an extra five or ten percent performance. So we weren’t taking too much of a hit in making something a little bit smaller that we could actually build. But, again, the copter still ends up being absolutely huge.

What conventional designs did you reference when you built your human-powered helicopter?
The conventional tail rotor configuration. The tail rotor basically wastes power. What we were doing at all times was looking for the minimum power configuration because, basically, the pilot’s power output is fixed.

We looked at counter-rotating coaxial rotor setup, though we ended up going with the quad-rotor configuration. We also looked at what’s called a reaction drive system, which was basically a single main rotor, but instead of driving the rotor at the root (at the center with a shaft) there were wings on the tips of each rotor [which we removed from the final design to reduce weight].

Why did you settle on quad rotors?
What all these configurations have in common is that they are all torqueless. You need to have a drive method such as that when the pilot drives the rotors, the body of the helicopter doesn’t just spin. You can do this with an even number of multiple rotors: two, four, six, eight would all be torqueless. Or you need to have a tail rotor to counteract torque, or you need to have a reaction drive.

What’s the maximum the pilot could weigh?
It was designed for a pilot weighing 165 lbs.

Could the copter ever fly outside?
Human-powered aircraft are incredibly light and incredibly large. They want to blow away in the slightest breeze. Even winds of as little as 3 kph, or 2-3 knots, would destroy the aircraft. Because we can only tolerate low winds, we would only be able to test outside every twenty days or so. It is even tough to find a place to fly indoors. An aircraft hangar built to accommodate a 747 is too small for comfortable flying.

What companies were the primary equipment suppliers? At the center is an off-the-shelf bike frame from Cervélo. It’s the lightest in the world. We designed and fabricated the mechanicals ourselves. The truss, the large structure that suspends the pilot and connects the rotors, is constructed from carbon fiber tubes we made by hand. They are incredibly thin and there is nothing else like them in the world.

Was the pilot ever in any danger?
In human-powered aircraft, the rule of thumb is you never fly higher than you’re willing to fall. We were very fortunate to be flying inside a soccer center because the field is like Astroturf. It’s also quite long and it’s got a rubber substrate in the under pad, so it’s quite bouncy. The pilot only got bruised once, never anything worse than that. And even in the two largest crashes from about 10 feet, he basically fell out of the sky, and in both he was unharmed.

The quad-rotor design of the Atlas met the standard requirements set by the American Helicopter Society, reaching 3-m altitude and flying within a 10-m border.
The Cervélo bike frame is suspended by Vectran wire bracing, which connects to a carbon fiber truss structure. The pilot generates 550 W. of energy to power Atlas, up to 60 seconds.
Here is one out of four rotor spools, connected under the rotor, to a drive spool on the bike frame. When the pilot pedals, he pulls the Vectran drive lines.

Voice your opinion!

To join the conversation, and become an exclusive member of Machine Design, create an account today!