Paul Dewhurst and Dr. Paul Robertson, chronicled in an earlier entry about their simultaneous electric ultralight flights last year, have created another first, a parallel hybrid motorglider launch. The truncated flight of their Alatus took place on September 9, 2010 at Sywell Aerodrome, between Coventry and Cambridge north of London.
Dewhurst notes, “Flight was quite short though (around 1.5 minutes) after the controller suffered a bit of tantrum complete with sparks! Rework [is] in process and we hope to have a rather longer flight soon.”
Replacing the Lynch unit normally used in the Alatus’s motorized version with an unspecified 76cc, 2.8 kilowatt (3.75 horsepower) internal combustion four-stroke leaf blower unit paralleled with a Hacker A200 12 kilowatt (16 horsepower) large model airplane motor approximates the Lynch motor’s output. Despite that, the airplane seems reluctant to leave the ground in the video.
Dewhurst explains, “Ground roll is quite long on the film, not entirely due to low thrust – it was uphill slightly and grass had just been cut and was just the right consistency to roll up behind the wheel and bunch up in the wheel box. I think on a hard surface we should be off in around 100-150 meters (320-480 feet).”
Both engines are connected through their output shafts, turning together to drive the single propeller. According to Dewhurst, “The IC motor should almost have enough power to keep the Alatus in level flight at 80-percent, so will need just a little of the electric power to help it. In theory this should mean we can get around 1.5 hours duration using around 1 litre (.26 gallon) an hour and running the 2.4Kw cell pack down to 20 percent or so minimum. But that’s theory at present!”
The Kokam lithium-polymer battery pack consists of 16 cells weighing 18 kilograms (39.6 pounds) with a total capacity of 2.4 kilowatt hours.
“As always, the installation wasn’t without its challenges to get this far, we ended up rewinding the hacker after it burnt out during ground running at a rather lower temp than we expected. Paul [Robertson] has rewound it with his special take on heat transference to the core, using finer filaments and high temp potting agents and vacuum bagging to squeeze it all tightly together.”
The Hacker develops at least as much power as in its original configuration, operating temperatures seem lower, and maximum temperature limits should be much higher. Dewhurst states that “Taking the IC motor out of its leaf blower home has required some work to get to cool satisfactorily too.”
The relatively low cost of the components for this conversion should appeal to other experimenters. 20 horsepower from a 26 kilogram (57.2 pound) package is as good as most four-stroke engine packages of that power, while offering radically better fuel economy. As a practical matter, any airplane that can take off and fly on one quart of fuel per hour while seeking out free thermals will have a place in green aviation’s future.
Since there is no specific reference to clutch arrangement to prevent it, I would point out that one possible inefficiency matching the symptoms described is that the engine is periodically opposed by the motor. A brushless motor of the type described chases a timed pulse. Making it go a different speed than it is presently slaved to electrically can result in it retarding the driving engine after current drops to nominal level (signaling zero delivered torque). Subsequent overspeed raises current again as the motor fights back, which could be a source of unexpected motor heating. Meanwhile, back EMF from a driven motor is often responsible for controller failure. Making the two systems cooperate is quite a challenge. Thanks for taking it on!
You are correct John in assuming there is no clutch between the engine & motor – it would be nice for a variety of operational reasons to have one, but for now, complexity & weight preclude it. As for the motor, we have actually included Hall position sensors to determine the phase drive commutation – mainly because of the high static torque load from the IC engine on start-up, but it does also ensure correct phase driving under +/- torques. In fact we have used the motor in generator mode as a dynamometer for the IC engine (minus prop.) to verify the engine power. I didn’t think the conventional ‘twitch it and see’ method for sensorless BLDC motor start-up would work in this application, so I designed and built a custom controller. The failure was in one phase of the output driver transistors – likely due to inductive switching transients exceeding the voltage rating of the MOSFETs. We did a number of ground runs with no problems, but with some airspeed the prop spins faster – giving a higher back emf, which together with freshly charged cells, probably popped us over the limit. I am now fitting higher voltage rated MOSFETs and paying more attention to voltage transients ! The thermal issue for the motor is now much more manageable than originally – when I rewound it I got 40% more copper into the stators, plus the vacuum impregnation of thermally conductive resin, gives a very good result. You make an astute comment about system cooperation – we have yet to grapple with how best to combine the ‘throttle’ settings of the 2 drives. Compared to our all-electric Lazair predecessor, this one is certainly “complicated” !
And here is some very interesting news
in the most unlikely form.
may be offering an enticing range extending formula for larger electric hybrid aircraft. It all hinges on weight and fuel consumption but the jaguar microturbine alternator set could be the bridge to allow the design of heavier vehicles ahead of the development of higher energy density batteries or functional super capacitors.
In case you have not considered these, a sprag clutch between the motor and the output shaft might not add significant weight or complexity, Paul R.