AT THE front of the room sits a giant moth, its massive 1-metre wingspan
displayed to full effect. The creature鈥檚 exoskeleton has been peeled away so
that the small audience can glimpse what lies inside. There are no guts and
sinewy muscles in here. Instead, the voluminous body is filled with intricate
networks of motors, gears and levers. At the flick of a switch, the insect
sparks into motion.
Slowly, the moth鈥檚 wings begin to trace a path of extraordinary complexity.
In one continuous movement, they sweep down and forwards, rotating about their
long axes and then up and back. Even the camber of the wings changes during each
beat. Charles Ellington has christened his mechanical moth 鈥渢he flapper鈥�, a name
which belies the sophistication of a machine built to mimic the wing motion of a
real insect with unprecedented accuracy.
Ellington, a zoologist from the University of Cambridge, is showing off his
invention to an international group of scientists who met in August at the
university鈥檚 zoology department. They are an eclectic bunch鈥攂iologists
mingle with engineers, materials scientists and robotics experts. Yet they are
united by a common interest in how insects fly, and a sense of renewed optimism.
After a long spell in the doldrums, the study of insect flight has fresh wind in
its sails. Ellington and his colleagues around the world have finally begun to
unravel the mysteries of insect flight.
Advertisement
鈥淎bout five years ago, insects couldn鈥檛 fly鈥攏ot according to the
conventional laws of aerodynamics,鈥� says Ellington. Analyses of insects in
flight could find only about one-half to one-third of the lift needed to support
their weight. Similar calculations showed that small birds and bats also
conjured up mysterious and unknown sources of lift.
Wherever the lift comes from, these tiny fliers apply it to great effect.
Insects can fly forwards and backwards. They can also hover and manoeuvre with
greater agility than the most advanced jet fighter. So much for the laws of
aerodynamics.
The problem is this. Conventional aerodynamics鈥攗sed in the design of
aircraft and helicopters鈥攔ely on 鈥渟teady-state鈥� situations such as a fixed
wing moving at a constant speed or a propeller rotating at a constant rate. By
contrast, the motion of insect wings is a complicated 3D affair. Nevertheless,
until recently researchers were not convinced that this special motion could
generate any unusual sources of lift. For years, they struggled to explain
insect flight using a theory rooted in steady-state situations, not
understanding why their aerodynamic sums didn鈥檛 add up. Ellington summarises it
neatly. 鈥淪ince the 1950s, we鈥檝e been looking at insect flight with the wrong
picture in mind.鈥�
This picture left out some obvious differences between insects and aircraft.
For a start, insects are small. On this smaller scale, the viscosity of air
becomes more important so that, for the average insect, flying through air is
like swimming through treacle. Because of this, the classic aerofoil shape that
generates an aircraft鈥檚 lift does not work, and insects have evolved entirely
different forms of wing.
Strangely shaped
At the University of Tokyo, Keiji Kawachi has been studying the conventional
aerodynamic properties of small, insect-sized wings as part of the Kawachi
Millibioflight Project (see 鈥淏reaking the laws of flight鈥�, New
快猫短视频, 18 November 1995). For medium to large sized insects such as
butterflies and dragonflies which have wingspans of between 5 and 10
centimetres, thin plate-like wings with a slight camber produce the most lift.
But Kawachi has found that even thin membranous wings are too cumbersome for
smaller insects to drag around, and nature has had to think again. The tiny
thrips, an insect with a wingspan of only a millimetre or two, cruises the
aerial highways on wings which are little more than hairy stalks. And does it
extremely well.
So how can these strangely shaped wings create lift? Whether attached to a
Boeing 747 or a butterfly, a wing can produce lift only if relatively fast
moving, low pressure air flows over its upper surface. Lift is generated by the
pressure gradient set up between the lower and upper surfaces of the wing.
For conventional aircraft, the wing鈥檚 camber and its angle of attack (its
incline relative to the airflow) create the necessary low pressure air over the
upper surface. But the lift force created in this way just isn鈥檛 enough to keep
an insect in the air. 鈥淪omehow, insects produce two to three times more lift
than you鈥檇 expect,鈥� says Ellington. The only way to explain insect flight would
be to find another way of producing low pressure on top of the wings to augment
the lift generated by conventional mechanisms.
The way paper aeroplanes fly offers some clues. Just before a paper plane
comes to rest, when the tail begins to drop and the nose sits high, it gains an
extra bit of lift before swinging lazily back to the ground. This phenomenon is
known as delayed stall and occurs when sharp-edged wings cut through the air at
high angles of attack.
To understand what is going on, imagine the air striking the leading edge. At
low angles of attack the air flows smoothly over the surface creating the
pressure difference that leads to lift. But at high angles, the flow breaks away
from the surface and begins to turn somersaults, forming a vortex. This vortex
sticks to the upper front portion of the wing, and the swirling, fast-moving air
creates a low pressure area that generates lift. It is this leading edge vortex
(LEV) that gives a paper plane its final boost before it lands.
But delayed stall does not keep the plane airborne for long because the
effect is only momentary. The rotating air cannot spiral in on itself forever,
so the LEV very quickly becomes unstable and tumbles away from the wing鈥檚
surface.
Nevertheless, Ellington and his team recognised that these kinds of
aerodynamic effects could play an important part in insect flight. The next step was
to work out how they were created during the complex beating of an insect
wing.
Monitoring and interpreting the 3D airflow across small flapping wings is not
easy. As well as seeing the motion of the airflow, researchers also have to
measure its speed and direction. To start with, Ellington鈥檚 team chose to work
on the hawkmoth, Manduca sexta,because of its large, 10-centimetre
wingspan, and because its wing motion is typical of many other insects.
Manducabeats its wings 26 times a second and by tethering the moth in a
wind tunnel, they monitored the path of horizontal smoke trails flowing over the
insect鈥檚 flapping wings.
鈥淲hen we got the flow visualisation results from Manduca we didn鈥檛
know what to make of it,鈥� says Ellington. As expected, his team could see a
vortex forming along the leading edge of the wing. But rather than tumbling off
the back of the wing, the vortex clung to the surface while the air within it
seems to spiral out towards the wing tip like a whirlwind.
Larger than life
Exactly what caused the spiral flow they couldn鈥檛 say. 鈥淲hat we needed was a
bigger insect so that we could see what was going on in more detail,鈥� says
Ellington. There are few insects in nature bigger than a hawkmoth, so in 1994 he
and his colleagues decided to build their own.
One of the peculiar predictions of aerodynamic theories is that it is
possible to mimic fast flow over a small object by generating a slow moving
airflow over a large object. Aerospace engineers reverse this principle to test
scale models of aircraft in wind tunnels. All Ellington and his team had to do
was build a model of the hawkmoth ten times larger than life. To create the
required pattern of flow, this model would have to flap its wings at one
hundredth the frequency of the real thing.
After nine months of design and construction, the flapper was born at a cost
of 拢40 000. And although its slow motion ensured that the flapper would
never get airborne, it was perfect for visualising the detailed airflow over the
wings.
Back in the Cambridge conference room, Ellington points to the crucial
evidence that he gathered last year. A video flickers into life, showing the
flapper. The head-on view of the model shows the wings raised, with their
tips pointing almost vertically upwards. As the down stroke begins, small plumes
of smoke emerge from within the body of the wings at points along its leading
edge. The smoke detaches from the wing and begins to somersault into a spiral.
Clearly, a leading edge vortex has formed.
Then something astonishing happens. During the down stroke, the centre of the
spiralling smoke is pulled out along the leading edge of the wing like a party
streamer. This enables the whole vortex to stay stuck like a limpet to the wing
surface until well beyond the halfway point in the down stroke.
This was just the evidence Ellington and his team had hoped for. With the
vortex stuck to the surface of the wing, they now had another lift-producing
mechanism. 鈥淥ur observation is a confirmation of what we knew had to be there,鈥�
says Ellington. At low air speeds, the vortex is extremely small and it takes
something as large as the flapper to see it properly. 鈥淭hat鈥檚 one of the reasons
why it wasn鈥檛 spotted in earlier experiments,鈥� he says.
Ellington calculated how much of an effect this might have. The results were
conclusive. The lift force generated during the downstroke was about one and a
half times that needed to lift the weight of Manduca鈥攎ore than
enough to keep it aloft. Last December he published his results in
Nature (vol 384, p 626).
Despite this momentous discovery, Ellington鈥檚 feet remain firmly on the
ground. 鈥淚t鈥檚 both the end of a chapter and the beginning of a new one,鈥� he
says. 鈥淲e鈥檝e observed a new aerodynamic phenomenon, but we don鈥檛 know how it鈥檚
produced.鈥� Nobody knows how the helical flow of air is maintained across the
wingspan. One possibility is that because the tip of a flapping wing moves
faster than its base, an area of relatively low pressure air at the tip might
suck the vortex along the wing. But other observations show that air moves in
the opposite direction beneath the wing. Further work is needed, says
Ellington.
Kawachi, who recently detected a similar vortex moving along the wing of the
dragonfly, is equally pragmatic. 鈥淚t鈥檚 one step towards understanding the
complex phenomenon of insect flight,鈥� he says. But there are many mysteries that
remain unsolved. Ellington and Kawachi have detected signs of other vortices
generated during each beat but have yet to measure them accurately enough to
gauge their effects. And while most research has focused on large insects, there
is no guarantee that the same high-lift mechanisms are at work in much smaller
insects such as the thrips. Then there is the question of the number of wings.
Many insects have four of them but just why this might be better than two is
still a mystery.
Despite these unanswered questions, engineers are already hoping to
capitalise on Ellington鈥檚 discovery to make tiny artificial machines. 鈥淚nsects
are ideal models for microrobots,鈥� says Isao Shimoyama. 鈥淭heir flight mechanisms
have already been tried and tested through millions of years of natural
selection.鈥� Shimoyama is from the University of Tokyo where he spends his time
designing and building tiny machines. He shows a picture of one of his early
models鈥攁 mechanical vehicle that is dwarfed by what looks like a large
metal hoop that on closer inspection turns out to be the eye of a needle.
Microrobot technology is still in its infancy, but Kawachi has no doubts
about the impact it will have. 鈥淚 believe that miniaturised flying robots will
have a big effect on human life,鈥� he says. Kawachi鈥檚 vision includes microrobots
that shuttle drugs to specific sites within the body or perhaps carry out
repairs. He believes it might even be possible to build microrobots from the
same material that insects themselves are made from. That way, the material
would break down inside the body after the mission was over.
The US Air Force is more interested in microrobots for surveillance purposes.
鈥淭hey want to build something that will flap like an insect, for disguise, and
they want it to fly at over 100 miles per hour,鈥� says Ellington
(see 鈥淧almtop planes鈥�, 快猫短视频, 5 April 1997, p 36).
In twenty years鈥� time,
that annoying black blob circling the living room lampshade above your head
might deserve a second look.
For the moment, though, Ellington, Kawachi and others are pleased with their
progress. 鈥淲e are teaching the aerodynamicists something new. And that鈥檚 a nice
feeling,鈥� says Ellington.
