Flight insights from flies

How does a fly fly and why should we care? To the first, says Michael Dickinson, professor of bioengineering at Caltech, the short answer is different from what we have thought. He and his colleagues used a dynamically scaled flapping robot (aka Robofly), a free-flight arena (Fly-O-Rama), and a 3-D, infrared visual flight simulator (Fly-O-Vision) to prove it.

And we should care, says Dickinson, because the simple motion of a fly flying links a series of fundamental and complex processes within the physical and biological sciences. Studying a fly may eventually lead to a model that will provide insight into the behavior and robustness of complex systems in general, and may help in the design of flying robots that mimic nature.

In a paper, “The Aerodynamics of Free-Flight Maneuvers in Drosophila,” Steven Fry of the University of Zurich, Caltech research assistant Rosalyn Sayaman, and Dickinson show how fruit flies use their wings to generate enough torque to overcome inertia, not—as conventional wisdom has held—to overcome friction. The paper appeared in the April 18 issue of Science.

Flies are capable of making 90-degree turns, called saccades, in less than one 50-thousandth of a second. To make the turn, a fly must generate enough torque, or twisting force, to offset two forces working against it—the inertia of its own body and the viscous friction of air.

Until now, it’s always been assumed that viscosity, or resistance to flow, is the enemy of small flying critters, while inertia is the bane of larger animals like birds. But the theory had never been tested.

To study the aerodynamics of active flight maneuvers, the researchers employed three infrared, high-speed video cameras (the 3-D Fly-O-Vision) to capture fruit flies, Drosophila melanogaster, performing saccades in free flight. The animals were released in a large enclosed arena (the Fly-O-Rama) and lured toward a cylinder laced with a drop of vinegar. As the flies approached the cylinder, it loomed within their field of view, triggering rapid turns that helped the flies avoid a collision.

The flies performed saccades within the intersecting fields of view of the three cameras, which allowed the researchers to film the turn, measure the wing and body positions, and calculate the velocity of the fly’s path.

The 3-D video of these saccades showed that, despite the fly’s size and slow speed, it typically performed a banked turn, first accelerating then slowing as it changed heading, then accelerating again at the end of the turn. This suggests that the timing and velocity of the fly’s turn are dominated by body inertia and not friction.

To see if the measured patterns of wing motion were sufficient to explain the saccades, the researchers played the sequences through a robotic model (Robofly) to measure the aerodynamic forces over time. They found that the timing and torque calculations based on the videotaped fly’s morphology and motion matched the calculations derived from the robot’s wing motion. These results further support the notion that even in small insects the torques created by the wings act primarily to overcome inertia and not friction.

Although the experiments were performed on fruit flies, the importance of inertia over friction increases with an animal’s size; thus, these forces impact nearly all insects. The results provide a basis for future research on the neural and mechanical basis of insect flight and, for roboticists, may offer insights into the design of biologically inspired flying devices.
•