Model Behavior

In the age of computer design, why do engineers still send airplane models to the wind tunnel?

Before supercomputers, wind tunnels quantified performance. Reference measurements on this model are used to determine the cross-sectional area for tests of a modified F-8's supercritical wing. (NASA)
Air & Space Magazine | Subscribe

A dust-colored Dodge Dart station wagon races down a narrow road in the Mojave Desert, north of Los Angeles. Six feet above the windshield is a model airplane that looks as if it’s flying backward. Inside the car, the driver maintains speed by reference to a primitive air pressure meter suction-cupped to the windshield; another man, his bushy sideburns fluttering in the hot breeze, flies the airplane with a radio-control modeler’s control box and notes the voltages from crude force gauges aboard the model.

From This Story

It’s 1975. The man in the right seat is Burt Rutan, the model airplane is the VariEze, and the rig on the roof is his “car-top wind tunnel.” In the years to come, the VariEze will revolutionize airplane homebuilding and Rutan will become the enfant terrible of a generation of aeronautical engineers, masterminding the first private manned flight into space.

Rutan tested that model, and others since then, to validate the novel configurations that are his trademark. Immersing physical models in genuine wind has always been the most direct way to collect aerodynamic information and verify new insights. German aviation pioneer Otto Lilienthal and the Wright brothers conducted hundreds of model tests. Even the great engineer Gustave Eiffel, of Eiffel Tower fame, built large wind tunnels to test wings and whole airplanes well before the start of World War I. Only by viewing and measuring the behavior of scale models in wind tunnels, and by using various tricks to reveal precisely what the air was doing, could aerodynamicists find their way to the best designs. The thick airfoils that made cantilever wings possible, the National Advisory Committee for Aeronautics cowlings that reduced the drag of radial engines, the fillets and fairings that doubled cruising speeds during the 1920s—all came from wind tunnel testing of models and full-scale airplanes.

Gerald Landry, who managed the California Institute of Technology’s famous GALCIT tunnel—Guggenheim Aeronautical Laboratories, California Institute of Technology—until its demolition in 1997, recalls the interactive approach to tunnel testing used by manufacturers like Northrop, Lockheed, and Douglas, which would contract with Caltech for tunnel time. “They would send a whole team, not just aerodynamics guys but the engineering and model shop guys too,” Landry says. “We’d run a test, and if there was a bad area somewhere, the shop guys would take material away or build it up and we’d run it again. I have my thumbprint in a number of airplanes’ wing root fairings.”

This was the procedure when the wing root and landing gear fairings of the six-passenger Northrop Alpha, an early TWA airliner and GALCIT’s first customer, were developed by trial and error in 1931. The Alpha was a triumph of empirical aerodynamics; the drag of the wing-fuselage combination was eventually reduced by half. The original wind tunnel model still exists; it is on display at the Western Museum of Flight in Torrance, California. But it is the exception. Countless of these statues of airplanes-to-be, although prized by collectors as works of art, were discarded or destroyed once they had yielded the necessary information.

Vast quantities of information emerged from decades of wind tunnels tests, filling libraries with reports on everything from the behavior of overall aircraft configurations to the tiniest details of structure or shape. The information found its way into multi-volume compilations of mathematical methods to predict the performance and behavior of new designs. But these methods were like systems to predict weather based on general observations of past trends, rather than on precisely extrapolating from present conditions in light of the basic laws of heat and fluid motion. As with weather forecasting, aerodynamics confronted a problem of sheer scale: The underlying physical laws had been known for a long time, but applying them to practical problems involved staggering numbers of calculations.

The advent of the fast digital computer made it possible to perform these calculations in a reasonable amount of time. Since the 1970s, computational fluid dynamics or CFD—the solution of aerodynamic problems by numerical simulation—has increasingly challenged the wind tunnel for the role of the airplane or missile designer’s most valued tool. Just as the proliferation of office computers inspired predictions of the disappearance of paper, the advent of CFD was claimed to herald the demise of wind tunnels. Both forecasts were premature. Wind tunnel testing still thrives, in part because it provides the indispensable final validation of computed results, and in part because for many kinds of tests it is actually cheaper and faster to build a physical model and test it than it is to prepare a complex and detailed computer model and then analyze it repeatedly over a wide range of flight conditions.

CFD has been, in a way, a perennial victim of its own successes. It is now called upon to solve problems that would have been considered impossible a decade ago. But the persistent obstacle for computer aerodynamics, according to Caltech aerodynamicist Mory Gharib, is the difference in size between the largest and the smallest objects it must deal with.

“Ideally, you would like to be able to analyze the smallest eddies in the flow, the so-called Kolmogorov eddies,” he explains. “But they’re very tiny. To examine the flow field surrounding a full-scale transport on this scale, you’d be calculating for years. So you have to be satisfied with looking at larger chunks, and that leaves you with some uncertainty. That’s where the wind tunnel comes in. It anchors your computational results.”

Cost, however, is a persistent issue. Engineers at Honda, enjoying an ample budget for development of a light jet, relied on a continual alternation of computational and wind tunnel results, obtained with a number of different models and in several different tunnels. They arrived at a design that is unconventional—the engines are mounted on pylons that sprout from the upper surface of the wing and the nose bulges strangely—but is claimed to be more efficient than the usual arrangement, in which the engines are attached alongside the rear fuselage. Sino Swearingen Aircraft Corporation, on the other hand, with a tighter budget for developing its own small twin-jet, bypassed wind tunnel testing of the wing, only to lose a prototype and its pilot in flight testing when an unanticipated transonic shock wave made the airplane uncontrollable in roll. Belated wind tunnel testing revealed the problem and provided the fix.

Comment on this Story

comments powered by Disqus