+ heli-sim
+ Roadmap
+ Scenarios
+ Download
+ Credits
+ Links
Mailing list:
+ Archives
+ Subscribe
+ Admin
+ PCB v2
+ PCB v1
+ Systems
+ Sensors
+ Field support
+ Prototypes
+ Summary
+ Bugs
+ Tasks
+ Stats
+ Non-SSL
<Blurry image of the first prototype in flight>

Design ideas

Here are some thoughts on how the cost goals were produced.

Onboard computer:			$ 300
Networking hardware:			$ 300
Servo controller:			$  50
Servos:					$ 200
AHRS:					$1300
Ground station (laptop):		$1500
GPS:					$ 100

Engine:					$1000
Airframe:				$2000

Onboard video equipment:		$ 200
Ground station video:			$ 150
Total:					$7100

[ Berger Industrial Twin ] Allowing for significant creep in the cost of the airframe, engine and AHRS still gives the project an easy margin to meet the $10,000 acquisition cost estimate. However, Chris Bergen from Bergen RC pointed out that these values may be far too low. His company produces the Industrial Twin helicopter (pictured on the left). It can lift 10 kg of paylod and costs $4,000. His estimates are that they can increase to 15 kg of payload and cost $6,000. To go to 20 kg would double the cost due to the larger/stronger components, rotors, etc. These numbers will have to be considered and our estimates revised.

VMax-6 engie In the prototype LM-110 (engine is pictured on the right), 1cc of engine displacement produces 100 watts and can lift 1 kg MGTW or roughly 10 kg/kw. Payload is a mere 200 grams. It has 607 mm main rotor diameter, which leads to a 3.4 kg/m^2 disk loading. If we scale this linearly to the full scale goals, the aircraft will mass 250 kg to be able to lift the target of 50 kg. To maintain the disk loading we need a main rotor diameter of 4.8 meters. However, an unofficial goal is to be able to store the UAV in an automobile or pickup truck. 250 kg is far too much for an ordinary person to heft into an automobile and the rotor diameter is far too large for even a long bed pickup. So clearly, a smaller rotor disk and less gross weight are required. Other computations show that linear scaling is too large in several parameters, so perhaps an empty weight is less than 100 kg is feasible. I do not have enough theory to compute a reasonable range of disk loadings.

The prototype uses a 1 cc (0.061 cubic inch) single cylinder two stroke engine that produces 100 watts. The linear scaling numbers suggest that a 250 cc engine would be sufficient, perhaps a single or two cylinder aircooled motorcycle engine. These can produce up to 75 kw, considerably more than linear scaling would estimate. A 1:2.6 gearing might be sufficient to keep the engine in its power band of 5000 rpm while maintaining a main rotor speed of 1900 rpm.

Fuel consumption of the prototype is 100 cc per hour. Again, if we scale from this basis, we can expect 25 liters / hour or about 6 gph. This would lead to a direct hourly cost of around $9 for fuel. Again, with a goal of $10/hour operational costs, this sounds reasonable. The estimated 250 cc engine would actually use considerably less fuel -- closer to 3 gph. For two hours of fuel, we have a high end weight of 32 kg and a low end of 16 kg.

Please remember, these numbers are being computed assuming linear scaling from a very lightweight prototype. Until more theory about the performance characteristics is known, these numbers are very rough.

Autopilot Logo SourceForge HTML 4.0!
$Id: design.html,v 1.33 2002/08/05 03:44:39 tramm Exp $