We wanted to send out a quick update on some of our improvements from this week. Here's what we've been working on:
Smoothing the Control System Response
We learned from testing this summer that we had plenty of authority with our control system. However, we knew there were some issues with imbalances in actuation (i.e. inconsistent response to control inputs, imbalanced response between rotor blades). We've been working to resolve a few of these issues, and have identified the source of most of our concerns. For example, we spent some time revising the control lines (stiffer, teflon-coated lines) and rotor tip pulleys, and have reduced the friction in these two components by 90%. More to come soon!
Throughout testing we typically rapidly-developed and jury-rigged solutions as necessary. A few of these solutions related to the truss, and we've been slowly making these improvised solutions more permanent and lightweight. As one example, we strengthened some of the bracing-line attachments, reducing the weight of that modification by 80%. Given that weight reduction is of primary importance for a human-powered helicopter, the importance of these modifications can't be overstated!
The canards have provided an important illustration of just how important weight reduction is, even versus aerodynamics. The canards are the all-moving control surfaces at each of the rotor tips. Since they're located at the tips they move extremely fast, and hence keeping them smooth and clean aerodynamically should be extremely important for reducing overall power. The method we had chosen to construct the canards initially did not lend itself to a clean result (wrinkles in the wing skin, deformed shape, etc). We knew this needed improvement.
This week we removed the mylar skin from the canards and made several improvements for comparison. First we replaced all of the trailing edges, formerly Kevlar lines meant to take the skin tension. To provide additional tension against our control lines, the canards had been supplemented with a curved plastic ramp at the trailing edge (called a “reflexed trailing edge”) that generated an additional aerodynamic force. However, this proved to be extremely draggy. Therefore, we removed the Kevlar line and reflexed trailing edge and replaced it with a triangular balsa wood trailing edge, of the kind typically used in model aircraft. This provides a much better shape and better bears skin tension.
We also knew that the leading-edge of the canards was likely bumpy and rough, resulting in turbulent flow over the entire surface and hence more drag. We tried two methods of cleaning this up, first simply sanding the foam leading edge and second coating this foam surface with paper in order to provide as smooth a surface as possible.
You can see the difference above in the cleanliness of the two surfaces. This visible improvement took several hours of work, but the testing numbers below bear out the result. This highlights the sometimes marginal gains from aerodynamics in this aircraft.
Power Required for flight by Modification:
Atlas As Flown (August): 620W
Optimized Rotor Angle Setting: 486.4W
Inverted canard (with reflexed trailing-edge removed): 470W
Improved Canards: 464W
Improved Canards (with paper leading edge): 463W
Despite this small additional improvement, at this point we have reduced the overall flight power required by 25%, impressive by any measure. In the new year we'll be working on further improvements including revised rotor tip fairings, cleaned up winglets, and changes to the drivetrain. Keep an eye out for more updates.
Happy Holidays and Happy New Year!