WIND TUNNELS OF NASA

 

Chapter 7 - The Post-Sputnik Renaissance of Aeronautics

The New NASA Aeronautical Wind Tunnels

 

 

The Langley V/STOL Wind Tunnel

 

[101] One would expect wind tunnel technology to become simpler when test speeds are reduced from 100 000 mph in the exotic atmospheres of other planets to a mere 100 mph in the Earth's familiar air. This would be true if one were testing, say, a Piper Cub, but V/STOL craft introduce a whole new set of problems for the wind tunnel designer. At the simplest level, there are two new enigmas to puzzle out: (1) how to keep the strong downwash from the fans or jets generating vertical lift from radically disturbing the airflow in the test section and (2) how to duplicate the airflow near the ground at low forward speeds.

The latter problem is also encountered when testing automobiles in wind tunnels. Neither the automobile nor the V/STOL aircraft encounters uniform vertical distributions of air velocity due to tunnel wall boundary layer effects. In automobile testing, a wind tunnel ground board moving at the speed of the free stream air pulls the boundary layer along with it, making the vertical velocity distribution uniform. The same strategy suffices in V/STOL wind tunnels.

In the case of flow disturbances created by downwash, the tunnel designer has two options. He can make the test section so large with respect to the model that the wall effects are negligible or he can build test section walls with variable openings to dampen the effects of downwash much like wall slots in transonic tunnels.

The benefits of large tunnel size were effectively exploited in the Ames 40 x 80 foot tunnel where full scale powered V/STOL aircraft were successfully....

 


V/STOL wind tunnel

 [102] The slotted wall, one of the Tinkertoy interchangeable wall sections that helped reduce wall interference in the V/STOL wind tunnel.

 ....tested. To provide similar V/STOL research capability on a more modest scale, the Langley low speed 7 x 10-foot tunnel was provided in 1956 with a roomy 17 foot square test section as a modification of the settling chamber. In 1968, a 9 x 15 foot test section (with model support and data acquisition system) was inserted in the 175 mph return leg of the Lewis 8 x 6-foot supersonic wind tunnel. This was NASA's first V/STOL facility devoted solely to propulsion system integration. This modification of an already existing tunnel proved invaluable in helping integrate propulsion systems into V/STOL aircraft. Meanwhile, a wind tunnel specially designed for and devoted completely to the investigation of V/STOL problems was under construction at Langley. This facility was brought on line in December 1970.

The Langley V/STOL tunnel was simple and low powered when compared to the power hungry titans built for the space program. A meager 8000 horsepower electric drive system was ample for the 230 mph speed desired. The test section was large-14.5 feet high by 21.75 feet wide. Special vanes were placed ahead of the tunnel fan to quickly cut off all air circulation for zero velocity tests. So far, the design and construction of such an unaspiring tunnel would seem child's play. The challenging design problems came with the test section walls and ground board. Unlike any tunnels built before, the V/STOL tunnel test section walls were built on the Tinkertoy principle. They could be changed from solid to slotted to semi open simply by interchanging wall sections. Proper wall selection could radically reduce the flow disturbances caused by the aircraft's downwash. The moving belt ground board was also new and unusual. Traveling at speeds up to 80 mph, the moving belt ground board was a massive structure riding on a wheeled dolly and transported into position on railroad tracks. The dollies and tracks also conveyed the test models into the tunnel. The models could be checked out and calibrated beforehand on their individual dollies in a huge model preparation shop. Once ready they could be trundled into the tunnel....

 


VTOL aircraft

[103] A model of a VTOL aircraft with tilting engines mounted over the moving ground board of the V/ STOL wind tunnel.

 

....test section. The sheer versatility of this tunnel attracted a large array of models ranging from helicopters to VTOL jet fighters to supersonic transports.

 

The Ludwieg Tube Tunnel at Marshall Space Flight Center

 

When NASA was formed it acquired Wernher von Braun's Army rocket group at Huntsville, Alabama. These launch vehicle experts formed the nucleus of NASA's Marshall Space Flight Center. Large rockets, like the Saturn 5, were their stock in trade. Nevertheless, the Marshall aerodynamicists did develop a special wind tunnel that had application to high performance aircraft as well. This tunnel, called a Ludwieg tube after the German who suggested it in 1955, was built to simulate the aerodynamic loads buffeting large launch vehicles as they rose through the atmosphere reaching speeds between Mach 0.2 and Mach 2.0. Coincidentally, the Reynolds numbers attained by the Ludwieg tube approached those encountered during cruise by full scale jet transports, such as the Boeing 707 and Douglas DC-8. The conventional NASA wind tunnels could not adequately duplicate these high Reynolds numbers.

The 32-inch Ludwieg tube built at Marshall Space Flight Center was hardly complex, being only a long tube of constant diameter capable of storing air at 50 atmospheres pressure. The model to be tested was positioned in a test section that was sealed within the tube by a downstream diaphragm-not upstream as in shock tubes. When this frangible diaphragm was ruptured, the air in the tube expanded, rushing past the model in the process. The run times were short, but for a half second or less the model was bathed in airflow that was constant in pressure and temperature and displayed very little turbulence.

The most significant characteristic of Marshall's Ludwieg tube was the high Reynolds number achieved-roughly three times that in conventional existing wind tunnels. This capability found immediate application in basic fluid dynamic research as well as the determination of aerodynamic forces acting on launch vehicles. Unfortunately, the Ludwieg tube had limited use in testing winged aircraft because of....

 

 


drawing of the Ludwieg tube

[104] Phantom drawing of the Ludwieg tube wind tunnel built at Marshall Space Flight Center for aerodynamic research on launch vehicles.

 

...the high stresses encountered and the consequent distortions of the models. For example, a model of a jet transport (18-inch wing span) could be distorted by 1 1/2 tons of lift force- an impossible load to withstand. Thus, for high Reynolds number testing of winged aircraft, some new facility approach was required.

 

A Dilemma Resolved by Cold Logic

 

In the 1950s, the slotted wall wind tunnel made it possible to simulate transonic flight-at least in terms of flight Mach number. Unfortunately, that most ubiquitous aerodynamic parameter, the Reynolds number, was not matched accurately. In fact, none of the transonic wind tunnels built up through the 1960s came within an order of magnitude of duplicating the true flight Reynolds numbers of transport aircraft. The reason was not hard to find: The models employed in transonic tests were too small. Since the Reynolds number is directly proportional to model length, those aerodynamic effects dependent on the Reynolds number were distorted in wind tunnel tests.

The penalty for poor simulation of the Reynolds number is best seen in the complex nature of transonic flow over an airfoil. In subsonic flight, up to about Mach 0.8, the air flowing over the upper surface of the airfoil accelerates to supersonic speeds and terminates in a shock wave standing almost vertically on the airfoil surface. At the base of the shock wave, the boundary layer of air thickens and pulls away from the surface, creating a broad wake of fluctuating flow. This region of separated flow changes the airfoil's lift, drag, pitching moment, and other flight parameters. At the low Reynolds numbers available in conventional transonic wind tunnels, the vertical shock wave sprouts forward farther on the airfoil than it would if the true Reynolds numbers prevailed. Consequently, the region of separated flow m the tests is larger than it should be and the measured flight performance of the model more pessimistic than need be. However, no one knew how to correct...

 


photograph of transonic flow

[105] Schlieren photograph of transonic flow over an airfoil. The nearly vertical shock wave is followed by boundary layer separation that adversely affects lift, drag, and other flight parameters.

 

...for the pessimism, and aircraft were overdesigned to be safe.

So pervasive are transonic conditions that any solution to poor Reynolds number simulation would have a far reaching impact. Military aircraft fight at transonic speeds, and subsonic transports cruise at shock limited Mach numbers. The tips of whirling helicopter rotor blades penetrate the transonic region. An ascending space launch vehicle encounters maximum dynamic pressure and buffeting in the transonic regime. For a reentering spacecraft, stability and control are most critical in this same speed range.

In searching for a solution, the mathematical makeup of the Reynolds number (applicable to either a gas....

 


Wind tunnel measurements

Wind tunnel measurements of shock wave induced flow separation are pessimistic when compared with actual flight data. The discrepancy is due to the inability of most wind tunnels to simulate high Reynolds numbers.

 

....or liquid) provides clues:

 

Reynolds number = (density x velocity x length)/ viscosity

 

A classic way to increase Reynolds number is to increase the air density by raising the tunnel pressure. However, this stratagem, when carried to extremes greatly increases the model loads, stresses, and deflections-as experienced with the Marshall Ludwieg Tube. Velocity cannot be changed arbitrarily, for the test Mach number must be maintained. In addition, model length must be kept small because of tunnel cost. (Drive power increases as the square of the tunnel dimension; the cost of the tunnel shell increases...

 


temperature chart

[106] At very low temperatures, the Reynolds number rises dramatically as the viscosity of the test gas drops. The chive power also decreases.

 

....as the cube of the tunnel dimension.) Gases other than air, such as freon, are suspect because experiments have shown that the positions of shock waves in other gases could vary substantially from those in air. There is only one adjustable parameter left: air viscosity. Nature cooperates by permitting air viscosity to be reduced by lowering air temperature. Better yet, a cold wind tunnel requires less drive power in addition to providing higher Reynolds numbers for the model in the test section.

 

Langley Pilot Cryogenic Tunnel

 

This double advantage of temperature reduction was recognized in 1945, but it was not until the early 1970s that wind tunnel engineers gave serious thought to going far down into cryogenic temperatures and operating a tunnel just above the liquefaction temperature of air. The so called cryogenic wind tunnel promised to solve at last the Reynolds number problem. By evaporating liquid nitrogen ( 320° F) directly into the tunnel stream, the test section temperature could be reduced from the usual 120° F to about 300° F. The Reynolds number would respond

 


transonic
cryogenic tunnel

drawing

The 13.5 inch pilot transonic cryogenic tunnel constructed at Langley is distinguished by its heavily insulated components. (Photo by Bruce Frisch, courtesy Astronautics and Aeronautics)

 

[107] by raising by a factor of 6. Power to drive the tunnel would be halved at the same time.

In the late summer of 1972, Langley decided to erect a pilot cryogenic transonic wind tunnel. The test section was octagonal in shape and 13.5 inches from face to face. The tunnel operated at 5 atmospheres pressure up to Mach 1.2. Built under a sense of great urgency, the first tunnel runs began in September 1973. Tests quickly proved that the evaporated liquid nitrogen maintained a surprisingly uniform temperature distribution. More important, during tunnel operation at wide ranges of pressures and temperatures, but at the same Mach numbers and Reynolds numbers, pressure distributions and shock locations on a test airfoil remained remarkably constant as predicted. In other words, the concept of a cryogenic transonic wind tunnel was sound.

The small (0.3 meter) pilot cryogenic tunnel turned out to be an important research tool in its own right. At first it helped define the limits of cryogenic operation by determining how cold the tunnel could be operated without exceeding liquefaction boundaries. Then it was turned over to the Space Shuttle Program, where it assessed Reynolds number effects on rocket nozzle hinge moments and base drag. The importance of duplicating true Reynolds numbers was emphasized when the base drag measurements proved that similar data from NASA's noncryogenic tunnels of lower Reynolds number were seriously in error.


PreviousIndexNext