The Langley full-scale wind tunnel, which dated back to the 1931 biplane era, would seem to be an unlikely candidate for new aerospace assignments. But its cavernous 30 x 60-foot test section and relatively low air velocities made it ideal for testing models of more modern aircraft in actual free flight, particularly stall characteristics.
The free-flight test arrangement was simple. A carefully balanced, lightweight model was actually flown in the tunnel. Of course, it remained stationary to observers but its forward speed was effectively that of the wind tunnel air. Operating propellers produced thrust or, for jets, a jet of high-pressure air supplied through a slack hose sufficed. Three pilots, one each for pitch, roll, and yaw control, sent signals through a slack power and control cable. By carefully observing the model's behavior under different conditions of flight, observers could spot weak points in the design
before the aircraft was too far into the extremely expensive development cycle. For example, poor stall performance might cause the model to roll violently and possibly enter a spin. It was much better for this to happen with the model than a piloted full-scale prototype. In a similar fashion, model testing of VTOL aircraft during the critical transition from hovering to cruising flight ironed out design deficiencies cheaply and safely. The piloting of VTOL craft was in fact so tricky that when full-scale versions were ready for prototype flight tests, one test pilot first "flew" the aircraft model in the full-scale tunnel to get a feel for the response of the aircraft to the controls during the critical transition from hovering to forward flight.
A second and unusual tunnel that NASA inherited was designed to explore the formidable and poorly understood area of aeroelasticity. Modern aircraft, especially wings, are obviously elastic. Jet transport wings droop toward the ground during taxiing; of...
...course, the wing droopiness disappears in normal flight because of the lift forces. But what happens to these flexible wings as the aircraft accelerates to high speeds? Will the wings start oscillating and be torn apart? Transonic aerodynamics further complicated an already complex aeroelastic problem-there was no clear answer to this question. Aircraft designers needed definitive wind tunnel tests to assure them that their thin-winged aircraft would not experience flutter under any anticipated flight conditions.
Flutter had been recognized as a problem for many years, but research was limited largely to the study of aircraft components. It was the Boeing Company, during the development of the radical swept-wing B-47 in the late 1940s, that first recognized the need to test dynamically and elastically scaled models of the complete aircraft. The model could not be rigidly supported as in the past; it had to have the freedom to move as a free body in response to the applied loads. This represented a formidable task for the wind tunnel designer.
In 1954 NACA began the difficult task of converting the Langley 19-Foot Pressure Tunnel for dynamic  testing of aircraft structures. The old circular test section was reduced to 16 x 16 feet, and slotted walls were added for transonic operation. A new 20 000-hp electric drive motor was installed; the tunnel designers knew that it would not come close to the desired speed of Mach 1.2 at required pressure levels. But they had an ace in the hole. They simply substituted freon for air. Freon, a fluorocarbon widely used in refrigerators, transmits sound at only half the velocity as air. A given Mach number and dynamic pressure could be attained with about one-half the power needed for an air-filled tunnel. Reynolds numbers also increased with freon-and that was advantageous. So, too, was the duplication of a key flutter parameter, Froudes number, which is used when gravity terms are involved in the equations.
In addition to the energy-efficient use of freon to make a slow tunnel appear faster (a concept borrowed from the Low-Turbulence Pressure Tunnel of the 1940s) the Transonic Dynamics Tunnel was provided with special oscillator vanes upstream of the test section to create controlled gusty air to simulate aircraft response to gusts. A new model support system was devised that freed the model so that it could pitch and plunge as the wings started oscillating in response to the fluctuating airstream.
Early in 1960, after 8 years of intensive design and calibration, the Langley Transonic Dynamics Tunnel, the world's first aeroelastic testing tunnel, was ready for its first occupant. Fate had already selected the model to be tested: the Lockheed Electra.
September 29, 1959, was warm and humid in Buffalo, Texas. At 11:07 p.m., Braniff Flight 542, an Electra turboprop, cruising lazily overhead, made a routine report. One minute later there was a blinding flash, a deafening roar, and the Electra crashed without survivors. Investigation of the accident revealed that the left wing had failed, leading to the general disintegration of the aircraft and a fire. There was no trace of metal fatigue-no inkling as to the cause of the catastrophe. The Electra was conservatively designed and had been thoroughly tested. It turned out not to be a fluke accident. On March 17, 1960, another Electra crashed at Tell City, Indiana. Its right wing was found 11000 feet from the crash site. It now seemed clear that violent flutter had torn the wings off the two craft. The critical question was, what had triggered the wing fluttering?
The new transonic dynamics tunnel had just been calibrated; a one-eighth scale model of the Electra, complete with rotating propellers, was quickly readied.....
....for testing. The elaborate Electra model could even simulate changing fuel loads and different enginemount structural characteristics. These properties had suddenly become important because a Lockheed engineer had suggested that the Electra had stimulated the catastrophic fluttering all by itself through the coupling of engine gyroscopic torques, propeller forces and moments, and the aerodynamic forces acting on the wings. The engineers had a term for it-propeller-whirl flutter.
Working with great urgency (130 Electras were still flying, though at reduced speeds), NASA, Lockheed, and Boeing personnel found first that the structural safety margins of the Electra far exceeded requirements. However, as they reduced the stiffness of the outboard engine mounts, the gyroscopic torques of the engine-propeller combination led to a wobbling motion with a frequency of 3 cycles per second. This frequency was identical to the natural flutter frequency of the wings. The catastrophic flutter stimulus had been found. The wrenching of the engine reinforced the wing oscillations until the wing fell off. The fatal resonance could build up and tear the plane apart in 30 seconds. No one could explain how the engine mounts might have been weakened-possibly during previous hard landings or violent storms-but the wind tunnel simulations fit the real accident situations perfectly. All the Electra engine mounts were strengthened and the aircraft has been operating successfully and safely ever since.
 The Electra was not the only aircraft with flutter problems to be tested in the transonic dynamics tunnel. The original C-141 military transport encountered severe tail flutter. The F-15 fighter's horizontal tail also fluttered, and the F- 16 fighter with external wing pods in some positions produced wing flutter. Modern aircraft are designed to bend under loads- but not too far-and it is the role of the wind tunnel to assure that the Electra story is never repeated.
The transonic dynamics tunnel, however, could not simulate structural problems at supersonic speeds. At supersonic speeds, thermal heating changed the situation. For example, a wing slicing through the atmosphere at Mach 3 might experience temperatures of 500° to 600° F on the thin metallic wing surfaces due to aerodynamic heating, while the sheltered heavy wing spars might run at only 100° F. Theoretical analysis suggested a dangerous decrease in wing stiffness that might alter the dynamics of the whole aircraft. Something akin to the Electra situation might recur at Mach 3 because of this nonuniform heating. Thermal simulation at high Mach numbers required a different kind of wind tunnel.
The new tunnel, the third NACA carry-over, had to duplicate Mach 3 flight conditions and be big enough to test large-scale models. A 9 x 6-foot tunnel seemed about right, but running it at Mach 3 at the required pressure and temperature would take about 1 million horsepower-a level of electrical power Langley could not hope to provide on a continuous basis. Therefore, the tunnel had to operate intermittently, drawing on stored energy. A big tank farm storing 130 000 cubic feet of air at 600 psi was sufficient for a run of a few minutes. To duplicate more closely the heating encountered by Mach 3 aircraft, the test section was preceded by a stainless steel heat exchanger fired by propane burners that heated the test section air to between 300° and 600° F. The result-an unusually noisy monster was named the 9 x 6-foot Thermal Structures Tunnel.
The thermal structures tunnel quickly ran into a brand new sort of problem. The aircraft designers wanted to measure the integrity of the model under simulated aerodynamic and thermal forces, but when the tunnel was turned on, a shock wave propagated down the nozzle and slammed into the model. Another shock wave jarred it from the opposite direction when the tunnel was shut down. To protect the rather fragile models from such heavy-handed treatment, temporary model shields had to be devised. A second approach was to remotely insert the model after the tunnel got up to Mach 3 speed and retract it before shutdown.
Noise was a perpetual problem with the thermal structures tunnel. Like the open-circuit tunnels at Lewis, it was a colossal bugle that set all the ducks on the adjacent marshland into scared flight. A long sound diffuser was added to muffle the roar. Nevertheless, so unpleasant was the downstream vicinity that an elaborate 5-minute sequence of warning signals was set up to warn personnel in the area.
Actually, some good was derived from the highintensity, low-frequency noise spewing out of this tunnel. The noise spectrum nicely simulated the roar emanating from large booster rocket engines. Various space vehicle structures, sensitive instruments, and astronaut communications systems (complete with astronauts) were tested in the tunnel's noise field.
It was only fitting that this facility, whose roar shook the Earth, met its end on September 30, 1971, when its 600-psi tank farm blew up. The debris filled the air, smashed several parked cars, but hurt no one. Some of the tank farm piping had failed because of metal fatigue. The thermal structures tunnel had done its job.