Rotor Trimming Optimization Yields 20% Power Savings!


This week we spent several days at The Soccer Centre optimizing the rotor and canard angle settings (Angle of Attack, in aerodynamics) in an effort to save power. Throughout trimming and flight testing in August, the rotors were set at or just below the angles for which they were designed in HeliCalc (our design and optimization program). However, given the various realities of constructing the rotors (fairings and adapters, the final airfoil shape, surface textures) we weren’t sure whether this was actually the best angle for the lowest possible power. With our extremely ambitious schedule we didn’t take the time to do the trimming studies required to determine the perfect angle settings, thinking we would have plenty of power margin. We’re finding out now how much energy could have been saved!

Todd on the drivetrain/bike frame driving the test rotor.

Todd on the drivetrain/bike frame driving the test rotor.

Our testing has been set up with only a single rotor, independent of the rest of the helicopter. The rotor was set up on the testing rig we manufactured a few weeks ago, and the drivetrain/bike frame secured on a stand outside of the rotor area. The rotor test rig includes a platform on which we mount ballast masses, simulating 1/4 the mass of the full pilot/helicopter. Todd drives the rotor by pedaling as would normally be done in flight, using the same Vectran drive lines. This method of testing (rather than using an electric motor to drive the rotor) accounts for drivetrain inefficiency in our power measurements, and also helps Todd practice a steady pedaling cadence and get used to the drive inconsistencies we’ve encountered in flight.

Rotor spool mounted on test stand with ballast masses.

Rotor spool mounted on test stand with ballast masses.

Our power measurements are taken using Look Keo2Power pedals, kindly loaned to us by La Bicicletta in Toronto. These pedals incorporate a force measurement system that determines cadence and force, therefore calculating power. We’ve tested these pedals against some of our other lab-quality power measurement systems, and we’re amazed by their accuracy! In consistent conditions and properly calibrated, we’ve found repeatability in our testing within 1-2 watts (roughly 1% for these tests). For a plug-and-play system brand new to market, this is incredible performance. With these pedals, we’ve collected data which will help save more than 100W!

Rotor airborne testing.

Rotor airborne testing.

Our optimization proceeded in two stages. First, we wanted to conduct a sweep of rotor angle settings to determine the minimum power rotor blade angle setting. In the plot below we see a clear minimum-power point at an angle of zero, where zero is actually our as-designed angle setting. This was reassuring news, as it effectively validates our design code. That being said, the power was higher than expected, indicating a higher proportion of profile drag than was designed for (profile drag arises from surface imperfections, inaccuracies in airfoil shape, protrusions and fairings, etc). The trend we found (as expected) was that profile drag reduced dramatically as the rotor speed was reduced. However at much lower speeds the angle setting to keep the rotor generating sufficient lift caused portions of the rotor to stall, increasing drag drastically.

Plot of rotor absorbed power versus rotor setting angle (without canard mounted).

Plot of rotor absorbed power versus rotor setting angle (without canard mounted).

The second stage involved a thorough analysis of the canard settings. The canards are used for control primarily, but do contribute to lift and drag. Initially they were designed to generate a downwards force, which created a “continuity of lift” from the root of the rotor to the tip of the canard. This simulates a longer rotor span and hence reduced induced drag (the portion of drag that arises simply from generating lift, and by far the greatest drag component for a helicopter). However, we knew that the downwards force would require additional power to fly the helicopter, but had not properly assessed this figure. For comparison we wanted to re-orient the canards to generate an upwards force also (creating more positive lift, thus reducing rotor speed and consequently profile drag). Lastly, we knew that a canard angled too far up or down would create excessive force and hence additional drag at the tip, and we wanted to find the neutral control angle with minimum drag.

Plot of rotor absorbed power versus canard setting angle.

Plot of rotor absorbed power versus canard setting angle.

What we found was in fact that an upwards-lifting canard required much-reduced power versus the downwards-force canard as originally designed. This savings alone (at the optimal rotor angle) amounts to about 37.5 watts.

Furthermore, we know that we’ll be able to save considerable additional power through further incremental refinements. Remaking the rotor tip fairings for the now-reversed canards, as well as improving the surface texture and airfoil shaping at the tips and canards could save an additional 10-15% on our overall power. Also, the up-facing canards allow us to remove the reflexed trailing-edge we had taped on to the canards to prevent the control lines from going slack under aerodynamic loads.

In conclusion our analysis from flight showed that the rotor setting of roughly -2 degrees from nominal and the downwards canard setting of 5 degrees required 170 watts of power per rotor. Now, with the rotor set at the nominal angle and the upwards-lifting canard at an angle of 2 degrees, the slower-spinning rotors only require 132.5 watts! That’s a 150 watt and 22% power saving over all four rotors, and we’re just getting started!

Comment