Frank Wenham operated the world's first wind tunnel in 1871. It took more than a century to advance to large, slotted wall tunnels with cryogenic cooling that make full scale testing at transonic speeds feasible and cost effective. Is there still room for improvement? While there are no quantum jumps in tunnel capabilities on the immediate horizon, several promising schemes for enhancing overall tunnel capabilities are well worth pursuing.
As long as we have had subsonic wind tunnels, the tunnel walls have distorted the flow of air around the models in the test sections. In most subsonic tests, the experimenter can make simple corrections for wall interference. In transonic and V/ STOL simulations, however, the corrections falter. Of course, one can always decrease the size of the model relative to the test section, but miniaturization of the model always compromises the accuracy of the tests, and complete, accurate simulation of aerodynamic reality once again escapes the experimenter.
The concept of an adjustable test section wall is not new. It is novel for a wind tunnel to automatically shape the contours of its test section to fit the streamlines surrounding the model. If an adaptive or self streamlining wall fits smoothly over the pattern of airflow, no wall disturbances will be created to propagate toward the model and upset the testing. The idea sounds good, but can it be accomplished? If the....
....wall can somehow "feel" the streamlines and adjust itself accordingly, the answer must be yes. Given the wall shape and the wall pressure distribution, aerodynamic theory can tell whether or not the tunnel wall conforms to a free air streamline. The wall contours are then adjusted to make the walls fit better, several times if necessary. It goes without saying that computers are heavily involved in calculating the degree of fit.
In both the United States and Europe, experiments with adaptive (or streamline) walls are progressing well, mainly on a two dimensional basis, although three dimensional trials will come soon. An encouraging feature of the early experiments is the discovery that a coarse adjustment of the walls can often reduce wall interference to the point at which the traditional mathematical corrections are once again adequate.
Unfortunately, the tunnel walls are not the only objects degrading measurements and distorting the...
....airflow through the test section. Even the Wrights realized that their measurements of forces in their primitive tunnels were grossly in error because of aerodynamic forces exerted on the supports holding the model in the airstream. These so called tare forces (from the Arabic ''tarha" meaning "deduction'') may exceed the forces on the model itself. The Wrights circumvented this unwelcome discovery by comparing test airfoils against a reference airfoil on a balance where tare forces canceled each other out. This is a good trick when only comparisons are wanted, but absolute aerodynamic forces must be measured if airfoil and aircraft performance is to be predicted. A search for better model support systems was initiated.
The early use of thin wires to support the model led to unforeseen disaster: The wind drag on the wires sometimes exceeded the model drag by a factor of 10. Streamlined support struts, shielded from the airstream by close fitting fairings, sent the tare drag plummeting. The resulting distortion of the airflow over the model, however, introduced a whole new universe of unwanted problems. In the early 1940s, the development of electric strain gage balances permitted experimenters to house the sensors directly within the model and then support it from the rear on a sting. This was a great improvement aerodynamically, even though the model had a bit of a bulge at the rear to accommodate the sting. It was better, but far from perfect, particularly at high angles of attack.
How can one support a model in a stream of air without any physical support? We cannot manipulate gravity, but we can generate powerful magnetic fields that will in effect negate gravity. In fact, the magnetic levitation of tracked vehicles has already become a reality. The magnetic suspension of models in wind tunnels is feasible, especially in the light of recent developments in superconductivity. Not only can magnetic fields support an aircraft model in a stream of air, but it is perfectly possible to measure aerodynamic forces magnetically as well. The magnetic lines of force can transmit the three components of force and the three moments exerted on an airframe to the supporting magnetic coils without intertering in any way with the flow of air. There is even the possibility that the model can be ''flown" magnetically; that is, the supporting magnetic fields can be varied to accelerate and maneuver the craft measuring the changing aerodynamic forces in the process. Magnetic suspension and magnetic balances seem almost too good to be true after decades of frustrating aerodynamic distortions from physical supports.
Scientists at the French ONERA first demonstrated the magnetic suspension of a model on a small wind tunnel in the late 1950s. Researchers at the Massachusetts Institute of Technology followed soon after.
Now aerodynamicists in many countries are experimenting with the concept.
Once the effects of test section walls and model supports have been either eliminated or compensated for, one might anticipate an aerodynamic nirvana. Nature and machines are not so kind. In wind tunnels, at least, one must still contend with the flow disturbances caused by the drive system and the flow channels leading up to the test section. There are three levels of air disturbance: (1) macroscopic eddies, swirls, and currents; (2) smaller scale turbulence within the airstream; and (3) molecular level noise propagated by sound waves. Ideally, all these disturbances should be rendered negligible before the. airstream reaches the test section. This, of course, never happens, but one can try.
The main battle with unsteady airflow is fought in the settling chamber upstream from the test section. Here the usual honeycombs and fine mesh screens strain out the random currents and vortices. The long stilling chamber muffles the discordance even more. As the airstream emerges from the stilling chamber in a good tunnel, the large scale disturbances have been attenuated to the point at which only the most sensitive flow measuring instrumentation can detect them.
Noise is a different phenomenon; it is propagated on the molecular level rather than the convective level. Noise slips through the honeycombs and screens preceding the test section with little attenuation, like talking through a screen door. Noise must be stopped at its source. Consequently, modern wind tunnel practice calls for installation of sound absorbing walls in the tunnel circuit near the major acoustic sources (the fans especially). If the noise peaks at certain frequencies, absorbing resonators tuned to the offending frequencies are very effective.
Progress in ironing out wrinkles in wind tunnel airflow, from large scale errant zephyrs to acoustic vibrations, has been slow but steady. An easily measured criterion of success is the distance along a surface air will travel before the boundary layer changes from laminar to turbulent flow. The better  the initial air quality, the farther the air flows along the surface before the crucial transition. The better wind tunnels can closely duplicate the air encountered in atmospheric flight.