TONY NEWTON / LONDON
Insect flight has provided the inspiration for research into a miniature air vehicle
At Cranfield University's Shrivenham campus at the Royal Military College of Science in Wiltshire, UK, Dr Rafa Zbikowski is taking inspiration from insect flight for a novel approach to indoor reconnaissance by tiny aerial vehicles the size of a hand.
In 1996, the Defense Advanced Research Projects Agency (DARPA), the US Department of Defense body which funds short-term, high-risk/high-return projects, had the idea of creating a Micro Air Vehicle (MAV) - a handheld, 150mm (6in) aircraft capable of reconnaissance and surveillance roles. Around $35 million was spent, largely on fixed-wing technology demonstrators, before the programme was shelved.
Currently receiving grants from the UK Engineering & Physical Sciences Research Council (EPSRC) and Ministry of Defence, Zbikowski's ideas differ from those of DARPA. "We have in common the concept of the MAV, but our thinking is very different," he says. "DARPA wanted an outdoor, over-the-hill reconnaissance vehicle. This may work in the Californian desert, but is not feasible in the wind and rain of Kosovo. There are many other assets which can fulfil outdoor reconnaissance, but there aren't any that can operate indoors, in confined spaces and in environments where current technologies just don't work - the sort of jobs that the military refer to as D3: dull, dirty and dangerous."
With 70% of the world's population living in urban environments, future conflicts are likely to be primarily urban. The urban environment is seen as a great equaliser: the lack of space and "round-the-corner" intelligence removes much of the advantage of Western military technology. It may be the military imperative which will make insect- like MAVs a reality, but civilian equivalents to D3 missions, such as assessing toxic environments, search-and-rescue and industrial monitoring, would also benefit.
The parameters of the Shrivenham MAV need to be different from those originally stipulated by DARPA. The MAV must be able to fly and be agile at low speed to avoid collisions with walls and other obstacles; to hover to allow precise observation; to have zero acoustic signature to allow silent penetration of hostile areas; to take-off and land vertically on unprepared ground; and to operate autonomously through intelligence or pre-programming.
Fixed-wing aircraft do not make good MAV candidates because, even scaled down, they need to maintain forward speed, can not turn tightly, fly backwards or hover, and find vertical take-off and landing (VTOL) difficult. They also have a power requirement of 150W/kg (68W/lb).
Miniature helicopters offer good low-speed agility, VTOL capability and can hover and fly backwards. However, they are noisy, have a high power requirement and suffer from recirculation problems that adversely affect lift and flight stability when operating close to a wall. In addition, debris sucked up by the rotor could be a problem in the environment in which the MAV would have to operate.
Having ruled out two options that have been around for less than 100 years, Zbikowski has turned to a technology that has been tried and tested by insects for 300 million years.
"If you want to flap like a bird with its bones, nerves and muscles, you have to create an internal skeleton, actuators and sensors. That means weight and tremendous complexity. But an insect wing comprises less than 1% of the mass of the creature's body, and is based on a spar-and- membrane construction, which contains sensors only for the purposes of detecting speed, direction and orientation. Motive power is supplied by mechanical resonance within the insect's body. It's possible to mimic that design with current materials and technology in a way that we couldn't with a bird's wing," says Zbikowski.
Manoeuvrability
"Insect flapping flight fulfils all the requirements of the MAV flight envelope. They can fly at low speeds - a typical insect flies at 7mph [11km/h], can hover, fly upside down, take off and land vertically, and are extremely manoeuvrable at low speed in all directions. They are inaudible if their wings beat at less than 20Hz, and experiments show that their power efficiency is typically 30W/kg, which is five times better than a fixed-wing/forward-thrust aircraft could achieve," he adds.
Zbikowski's aim is not to build a 150mm artificial insect, but to realise its functionality through engineering. The first task is to create a pair of wings which can reproduce the complex pattern of oscillation and rotation displayed by insects, and mimic the mechanical and surface characteristics of the wings that have been shown to contribute to lift, vortex generation and airflow.
In addition, the materials used have to be capable of withstanding hundreds of thousands of cycles of bending and twisting. The complex flapping mechanism involves a downstroke during which the leading edge actually reverses direction.
Zbikowski is working with Cambridge University entomologist Professor Charles Ellington on how the insect's anatomy relates to its flight capability, and to work out which structures must be created in order to mimic that ability, and which can be discarded.
"A conventional aircraft wing is smooth and rounded, whereas an insect wing is angular and has a rough surface texture," says Ellington. "Operating at low Reynolds number, in which viscous effects dominate, boundary layers grow and separate far sooner. Rather than trying to overcome this early transition, the insect exploits it. Corrugation near the leading edge provokes early separation, which gives rise to a separation bubble. The recirculation of this bubble effectively changes the shape of the wing to provide a thicker, smoother aerofoil to the incoming airflow and so gives a thick-wing profile while needing only the weight of a thin wing.
"What we don't know is whether these viscous effects will apply to a scaled-up version of an insect wing," adds Ellington. "But if we look at the fossil record, we find huge winged insects far bigger than the MAVs we are now trying to create. If they could fly, so should our MAV."
The fact this research is supported by AgustaWestland in the UK testifies to the seriousness of the endeavour in yielding useful data on unsteady and high-lift aerodynamics that could be relevant to helicopter design. By the end of the three-year EPSRC grant next year, the Shrivenham team expects to have produced a 150mm-long demonstrator, comprising wings and computer-controlled actuator.
At the Georgia Tech Research Institute in the USA a different approach is being taken with what they call the "entomopter". For Dr Robert Michelson and his team, the way forward is not the mimicking of insect flight mechanics, but the creation of a MAV which looks like an insect but doesn't really fly like one. "Control of a biomimetic wing requires adjustments in the angle of attack on the beat-to-beat basis, differential stroke extent, or differential stroke-rate between the two opposed wings.
"With the entomopter, we've addressed the issues of control and manufacturability from the very beginning. The entomopter wings beat autonomically - they don't need to beat harder on one side than the other to maintain stability or to navigate. This makes the wing-beating structure kinematically simple and therefore manufacturable. Stability and control is achieved through the beat-to-beat modification of the coefficient of lift of each wing by active flow-control techniques," he says.
Key differences
These techniques highlight another key difference between the two approaches. Whilst the Shrivenham team favours an electrical power supply, the Georgia State device is powered by a reciprocating chemical muscle (RCM), in which hydrogen peroxide (H2O2) is mixed with water in a catalytic chamber to create steam. The application is new, but the basic technology stems from the V2 rocket. Apart from providing thrust and directional control, the Georgia Tech team envisages that the RCM could also provide high-energy ultrasonic emissions for ranging applications, dry lubrication of moving parts, and active flow-control over aerodynamic surfaces.
Georgia Tech reports that it will soon test the fourth-generation RCM under contract to the US Air Force Research Lab's Revolutionary Technology Program. This unit will still be larger than the flyable version, but according to Michelson: "Reduction in size is largely a function of dollars, not new technologies." Georgia Tech's entomopter has secured $500,000 funding from the NASA Institute for Advanced Concepts to analyse and develop a scaled-up entomopter wing capable of operating in the lower Mars atmosphere as part of the Mars Surveyor project.
Back on Earth, Zbikowski is looking ahead to the control aspects of the programme and is working out what else can be learned from insects. The house-fly has a paired organ called the haltere on each side of the thorax. These halteres, which may be the vestigial remains of a second pair of wings, function as primitive gyroscopes and allow the insect to do what most cannot do: fly in the dark without visual cues.
Insect flight dynamics are complex, but their flight control must be simple. Zbikowski, whose background is in guidance and control systems, says: "We know that the fly's brain has about 500,000 neurons, and that 75% of these are engaged in vision processing. That leaves a maximum of 100,000 neurons for flight control - and the average toaster now has a bigger computer chip than that."
Source: Flight International