OLD wind tunnels never seem to die of themselves; they have to be killed, deliberately. There is a tendency to keep them going well beyond the point of diminishing returns. At Ames, the growing shortage of manpower made the abandonment of marginally producing wind tunnels an urgent necessity. The 1- by 3 1/2-foot tunnel was deactivated in 1954 and later completely dismantled. The old low-density and heat-transfer tunnels were abandoned, too, in 1954 when the new facilities were put into operation. The old heat-transfer tunnel was later consigned on indefinite loan (given) to the University of California at Berkeley.1 In September 1956 the 7- by 10-Foot Tunnel Section was disbanded and the tunnels turned over to the 40-by 80-Foot Tunnel Section to operate as needed.
Several tunnel modifications were also completed during this period. Beginning about September 1954, the 6- by 6-foot tunnel was shut down for major modifications that required nearly 1 1/2 years to complete. The modifications included measures to convert the tunnel for transonic operation. Toward this end, a slotted throat and a new model-support system were provided. Also, arrangements were made to inject air removed from the slotted throat into the supersonic diffuser, thus increasing diffuser efficiency and tunnel performance.
The 6- by 6-foot tunnel, as a result of the changes made at this time, could operate at Mach numbers ranging from 0.6 to 2.3. Moreover, the operation of the tunnel was made so simple that the task could easily be handled by one man. Indeed, one lonely night it was operated by a cat that somehow became entrapped in the switchgear. The regular operator was on hand finishing up his paperwork for the night when, to his surprise, he  heard the wheels begin to turn. Investigation revealed the cat-in what condition was never disclosed.
A number of relatively minor alterations were made in the 6-by-6 both before and after the major modifications just mentioned. The model-sup port system had earlier been rotated to accommodate an unplanned nonlinearity in the airflow, and the speed control of the tunnel had been altered to allow subsonic operation. Also, the aluminum compressor blades had given trouble and had been replaced by hollow welded steel blades. Cracking, which occurred in the steel blades as a result of vibration, had in turn been cured by filling the interior of the blades with a plastic foam. Following the major modification a fire, starting from an oil leak, ignited the magnesium compressor (J-33) blades of the auxiliary air system. The resulting soot, together with the water used to quench the fire, thoroughly fouled the whole tunnel interior.
A number of other tunnels at Ames underwent modifications during this period. The 14-foot transonic conversion of the 16-foot tunnel was completed in 1955; and general improvements, including the installation of more sighting stations and a Mach 3 nozzle, were made in the supersonic free-flight (SSFF) tunnel. Through the skillful efforts of Al Seiff and his staff, the great potential of the SSFF tunnel was being realized. Guns, instrumentation, and flow characteristics had continuously been improved and many ingenious research techniques exploiting the possibilities of the counterflow test principle had been developed.
The capabilities of the 1- by 3-foot tunnel had also, to a surprising degree, proved amenable to expansion. During the 1954-1955 period, in accordance with earlier plans, the power of the continuous-flow 1- by 3-foot tunnel was doubled and a new flexible throat, operated by hydraulic jacks, was installed. A year or two later a similar throat was installed in the blowdown tunnel. The new flexible throats reduced the Mach-number change time to a matter of seconds and with the further addition to each tunnel of a second throat-a constriction at the exit to the test section-the operating efficiency of the tunnels was greatly improved. The second throat increased the efficiency with which the supersonic stream of air could be slowed down, as it had to be, after passing through the test section. As a result of this benefit, the top Mach number of the power-augmented, continuous-flow tunnel was raised to 6.0. That of the blowdown tunnel was a little less than 6.0.
Among the major new wind tunnels completed during this period were the three Unitary Plan tunnels which went into operation in 1955. However, as earlier noted, the compressor of the 11- by 11-foot transonic leg suffered a blade failure in October 1956 with the result that the tunnel was out of commission until late 1957.
The power involved in running the equipment at Ames had reached large proportions by 1957. The total connected load amounted to nearly...
....half a million horsepower and the yearly power bill for electricity was approaching $2 million.
The interests of aeronautical research scientists had now advanced from the field of aerodynamics to the much more complex field of aerothermo dynamics. The heating to which high-speed aircraft would soon be subjected had become a matter of paramount importance to design engineers. A missile traveling at 7000 miles per hour, for example, could theoretically generate air temperatures behind its bow shock wave of as much as 8000° F- nearly as hot as the surface of the sun. The kinetic energy of the missile would be in effect converted to kinetic energy in the molecules of the surrounding air. The motion of the molecules thus heated would become so violent that they would break apart-"dissociate," as it was called. Thus a molecule of oxygen O2 would break down into two individual oxygen atoms. The process of dissociation would absorb heat, which would considerably alter the thermodynamic balance of the airflow around the missiles. The aerothermodynamic processes were complex and not readily amenable to theoretical treatment. Thus in this period a massive effort was begun to devise experimental test facilities in which the aerothermodynamic properties of hypervelocity flows could be simulated.
The idea of a ballistic range in which a test model could be launched into a body of still air by a gun was not new; in 1947 Ames engineers had proposed such a range to NACA management but the proposal had been rejected. However, the supersonic free-flight tunnel devised by Harvey Allen was in effect a rather sophisticated version of the ballistic range in which the air was in motion. In either case the gun was the critical item in determining performance. One of the fastest rifles built, the 220 Swift, developed muzzle velocities of about 4000 feet per second, and the ultimate velocity that seemed likely to be achievable with a rifle powered only by gunpowder appeared to be under 10,000 feet per second. It was clear that the flight of ballistic missiles which reentered the atmosphere at speeds of 20,000 feet per second or more could not be simulated by models launched with an ordinary gun.
As earlier mentioned,2 Alex Charters in 1952-1953 set about designing a high-performance "light-gas" gun in which the test model was to be propelled through a "launch tube" by a charge of highly compressed helium gas. The gas was compressed in a gun barrel or driver tube by a heavy metal piston driven down the tube by an exploding charge of gunpowder. The reaction to the explosion would be neutralized by driving a second piston out of the open rear end of the driver tube. In this manner, Charters and his two teammates, Pat Denardo and Vernon Rossow, hoped to achieve model launch speeds of well over 10,000 feet per second-perhaps as much as 20,000 feet per second. The maximum velocity actually achieved was about 13,000 feet per second. Charters and his colleagues did a thorough design job on the gun. They first developed the theory and then built several prototype versions. These piston-compressor guns are believed to be the first  light-gas guns built for model launching in this country and perhaps in the world. The gun and its development were described by Charters, Denardo, and Rossow in TN 4143 (ref. B-36).
An early small-scale version of the Charters gun was first incorporated in the Janus range in which the gun fired its tiny bulletlike model into a small, gas-filled cylindrical chamber 21 feet long. By 1955, however, plans had been laid for building a large range in which test models would be fired into a cylindrical test chamber 8 feet in diameter and 500 feet long. The test chamber, it was planned, could be filled with air or any other gas and could be pressurized or evacuated through a range of from 0.001 to 5.0 atmospheres. The construction of this range, called the main range, was approved by NACA management. Before going too far with its construction, however, the designers felt it desirable to build a prototype, much smaller than the main range but considerably larger than the Janus range. Construction of this facility, called the pilot range, got under way in 1956 as did also, a little later in the year, the construction of the main range. Robert Berggren and Paul Radach carried much of the responsibility for this work. Both ranges were completed in 1957. All three of the ranges used the piston-compressor light-gas guns and all were housed in a new building constructed along the western border of Ames territory on ground that had been reserved, but never used, for a seaplane towing basin.
Also proposed during 1955, and approved, was the construction of a pressure range in which the aerodynamic, the thermodynamic, and particularly the dynamic-stability characteristics of fairly good-sized models could be investigated at high Reynolds numbers and at Mach numbers of about 10. The pressure range, designed by Al Seiff and his staff, was put under construction in 1956 but was not completed until 1958. This range was located beside the 12-foot tunnel, and its test chamber, 10 feet in diameter and 200 feet long, was pressurized or evacuated by the facilities of the 12-foot tunnel. The range of pressures available in the new facility was from 0.01 to 10 atmospheres and, at the highest pressure, Reynolds numbers up to 300 million per foot of model length were obtainable. In all of the ranges, the test chamber was equipped with a row of sighting stations through which photographs, or shadowgraphs, of the test model could be obtained by electronically controlled spark or flash photography. The model was timed as it passed the sighting stations and drag coefficients were calculated from the rate of deceleration. In the pressure range, the observed pitching and plunging motions of the model were expected to provide data from which dynamic stability derivatives could be calculated if suitable analysis methods were developed.
Despite the tremendous effort put into the development of the pistoncompressor light-gas gun, the performance of the device proved rather unsatisfactory. It had a number of basic faults which in retrospect seemed rather obvious. Trouble arose from the heavy pistons. The idea of having the reaction piston fly out the rear end of the driver tube was really not practical; and the action of the other piston slamming against the end of the smaller diameter launch tube was notably destructive. The idea, of course, was to have a residual of gas stop the piston just before it hit the launch tube, but this was hard to arrange and introduced an element of thermodynamic inefficiency which reduced the performance of the gun.
Al Eggers and a number of other people at the Laboratory felt that it....
....should be possible to compress the driving gas in a light-gas gun with shock waves instead of with the heavy steel pistons used by Charters. The task of investigating this possibility was undertaken in 1956 by Carl Bioletti and Bernard Cunningham. They first built a gun, much lighter than Charters', consisting of a closed-breech driver tube and a smaller-diameter launch tube separated by a diaphragm intended to rupture at a prescribed pressure level. The driver tube held an explosive charge in its rear end but otherwise was filled with compressed helium. The exploding charge would generate a shock wave that would travel down the tube like a piston, heating the helium and compressing it beyond the rupture strength of the diaphragm barrier. As the diaphragm burst, the test model, mounted just ahead, would be hurled down and out of the launch tube. Carl and Bernie found that, while the shock wave did indeed serve as a piston, improved performance could be obtained if a very lightweight floating diaphragm was used to separate the helium from the gas generated by the exploding powder. In this way, more energy could be imparted to the test body, and muzzle velocities of 11,000 or 12,000 feet per second could be obtained with a gun much lighter and simpler than the one developed by the Charters team.
The next step in the development of what was now variously called the shock-driven, shock-heated, or shock-compression gun, was to install an intermediate shock tube, filled with helium, between the driver and the launching tube. In the operation of this two-stage shock-compression gun, the gas in the shock tube was compressed by a lightweight piston which was driven by the compressed gases of the driver tube. Proportioning of tube length and diameter as well as piston weight and powder charge were critical matters which Bioletti and Cunningham worked out with some care. The resulting gun, which was relatively simple and particularly adapted for launching lightweight models, was capable of achieving muzzle velocities of 20,000 feet per second. The two-stage gun and its development were later  described in TN D-307,3 entitled "A High-Velocity Gun Employing a Shock-Compressed Light Gas," by Carlton Bioletti and Bernard E. Cunningham.
Although a single-stage shock-compression gun was installed in the supersonic free-flight tunnel, the gun had been specifically designed for use in an interesting new simulator, called the atmosphere-entry simulator, or AES, which had been devised by Al Eggers. Al had become extremely interested in the heating problem encountered by missile warheads and other reentry bodies as they plunged back into the atmosphere. He had worked with Harvey Allen on the famous blunt-body concept and was all too aware of the limitations of theory in predicting the aerothermodynamic environment which reentry bodies encountered. Would it be possible, he wondered, to gun-launch a model into a test range in which the air density varied in the same pattern as that encountered by a body entering the atmosphere? It should be possible, he believed, to design a supersonic tunnel in which, by shaping the tunnel walls, the air density and pressure along its length could be made to vary in a prescribed fashion. Atmosphere entry might thus be simulated by launching a model into such a wind tunnel after the fashion of the SSFF tunnel. While it thus appeared that an atmosphere-entry simulator could be built, it would first be necessary to determine analytically whether a true simulation of reentry phenomena could actually be achieved in such a device. Eggers' analysis of this matter, together with a description of AES design features, is contained in TR 1378 (ref. B-37). Inasmuch as the analysis was favorable, immediate steps were taken to build the simulator. Al's analysis had been made early in 1955; a successful prototype was built in 1956; and the construction of a larger version, located in the range building, was begun in 1957. In the design and development of these unique facilities, Stanford Neice, of Eggers' staff, played a major role.
Al Eggers' work on reentry aerothermodynamics won much acclaim from the scientific community. For this work, he received, in 1956, the Arthur S. Flemming Award conferred by the Washington Junior Chamber of Commerce. This award is given each year to the 10 most outstanding young men in the Federal Government. Al, like a number of others at Ames, had earned his Ph. D. at Stanford while working at the Laboratory. He had a keen and very original mind, was a slashingly aggressive leader, and his ability to sell his ideas to others was of singular character.
The idea of using an explosive charge to operate a wind tunnel had....
....first been applied at Ames in the gun tunnel developed by Al Eggers. The gun tunnel, like Charters' gun, involved a heavy metal piston and did not Work very well. It led, however, to the development of the shock tunnel which used shock compression to generate a hot, hypersonic airflow of short but adequate duration. The shock-tunnel development proceeded in parallel with the gun development; in fact, one might be thought of as a hot-air gun and the other as a model-launching gun. In the shock tunnel, however,....
...it was found desirable to use a mixture of oxygen, hydrogen, and helium as the explosive element. A gas mixture of this kind, suitably proportioned, provided a burning rate and an impulse that were more compatible with the requirements of a shock tunnel than were those of a solid explosive.
Development of the shock tunnel in 1956-1957 was proceeding on two fronts, one carried forward by the team of Bernie Cunningham and Fred Hansen under the supervision of Al Eggers, and the other by Tom Canning who, under Al Seiff, was Assistant Head of the Supersonic Free Flight Tunnel Branch. Tom, following a suggestion earlier presented by Jack Stalder, proposed that the shock tunnel be used in a counterflow (a la SSFF tunnel) arrangement with a light-gas gun to achieve a relative velocity between model and airstream which would be much higher than ever obtained before. He sold Al Seiff on the idea and together they undertook to present it to Harvey Allen, their division chief. Harvey thought the idea was not very practicable. The expectation of being able to obtain useful data in the fleeting instant during which the pulse of air from the shock tunnel and the speeding model from the gun came together appeared overly optimistic and was almost too much for Harvey to swallow. Nevertheless, Tom and Al kept up the pressure and finally won Harvey's somewhat reluctant consent to try out the scheme. The trial was made with makeshift equipment, but was sufficiently promising to justify the construction of a pilot model of the arrangement.
The problem of naming the many hypersonic test devices that had been invented was by now becoming troublesome. The name chosen for the arrangement proposed by Canning and Seiff was ''hypervelocity free-flight facility," or HFFF for short. Thus what was under construction in 1957 was a relatively small and inexpensive pilot HFFF made up mainly  of parts scrounged from other places in the Laboratory and built, no doubt without Headquarters' knowledge or sanction. At the same time, Bernie Cunningham and his colleagues had designed and were beginning the construction of a 1-foot shock tunnel in which stationary models could be tested at very high airspeeds.
Although the ballistic range, the supersonic free-flight tunnel, and the hypervelocity free-flight facility all had outstanding capabilities for simulating the aerothermodynamic conditions of high-speed flight, they nevertheless possessed, in common, a couple of rather obvious faults. The conditions of model freedom and the short operating time rendered all measurements difficult and some impossible. Another problem was the severe restrictions on the size and complexity of test models. The models had to be small enough to be inserted into a rifle barrel and rugged enough to withstand accelerations of about 1 million times that of gravity.
In view of these circumstances, there remained a great need for facilities in which prolonged tests could be made at high Mach number and temperatures on fixed models of reasonable size. Hypersonic wind tunnels partially satisfied these requirements, but their speed capabilities were too low and the air-heating requirements troublesome. In such facilities it was necessary to heat the inlet air to fairly high temperatures to avoid liquefaction in the supersonic test section and to still higher temperatures if reasonably faithful simulation of aerodynamic heating effects was to be achieved. Indeed it could be calculated that the preheating requirements for simulating conditions encountered by a reentry body would lead to air temperatures of many thousands of degrees-well beyond levels that could be tolerated in a continuous-flow wind tunnel.
Though troublesome, the heating of wind-tunnel air seemed essential if aerodynamic-heating effects were to be properly simulated. The only question remaining was how such heating might best be accomplished. Shock heating had proved an effective means of increasing the temperature of the air in shock tunnels, while in continuous-flow tunnels, such as the 10- by 14-inch and the new 10- by 10-inch heat-transfer tunnel, preheating of the air had been accomplished by electrical-resistance heaters. In another preheating scheme being investigated at this time, the inlet air was passed through a thick bed of refractory pellets which previously had been heated to high temperature by a gas or an electric heater of some kind. It appeared that this device, called a "pebble-bed" heater, would serve for a short period of operation, perhaps a few minutes, but would not do for a continuous-flow tunnel. Late in 1957 Jack Stalder made an experimental installation of a pebble-bed heater in his 8-inch low-density tunnel. At the same time Al Eggers  and his staff were engaged in the design of a 3.5-foot-diameter hypersonic blowdown tunnel that would make use of a large pebble-bed heater.
The 3.5-foot tunnel represented a very ambitious project with an estimated cost of about $11 million. Its approval, first by Ames management and then, late in 1957, by NACA Headquarters and Congress, gave effective attestation to the salesmanship of Al Eggers. The intended purpose of the 3.5-foot tunnel was to investigate the aerodynamic and thermodynamic conditions encountered by hypersonic airplanes and boost-glide aircraft flying in the Mach-number range from 5 to 10. The tunnel would be a closed system. During its operation, vessels of highly compressed air would be released by suitable valving to pass through a pebble-bed heater, thence through the test section, and into four large evacuated spherical recovery tanks.
The 3.5-foot tunnel was to be equipped with a set of interchangeable fixed nozzles allowing operation at discrete Mach numbers of 5, 7, 10, and 14. Operating periods would last from 1 to 4 minutes. It was expected that the pebble-bed heater, a huge insulated steel pressure vessel filled with 125 tons of aluminum and zirconium oxide pebbles, would raise the inlet-air temperature to 4000° F. The new facility was to permit simulation of flight Reynolds numbers for Mach numbers up to 10 and of corresponding flight temperatures for Mach numbers up to about 7.0.
The tunnel walls, according to plan, would be cooled by a boundary layer of helium gas introduced through slots in the constricted, sonic throat and recovered downstream of the test section. The recovered helium gas would be contaminated with air, of course, and a purification plant would be required to process it for re-use. Indeed, the plans called for a separate, rather large building to house the impressive array of auxiliary equipment associated with the tunnel. Cooling of the tunnel was quite a problem. Aside from the helium cooling of the tunnel walls, the great steel shell of the pebble-bed heater, 8 inches thick and weighing over half a million pounds,4 would be insulated on the inside with refractory brick and be further protected by water cooling coils installed between the brick and the inner steel surface. Internal water cooling would be used to protect the model-support struts as well as the internal strain-gage balances with which the models would be equipped. The output of the balances was to be fed directly into an electronic computer for data workup. The 3.5-foot tunnel, it was realized, would be the most expensive piece of test equipment built at Ames since the construction of the Unitary Plan facility.
The pebble-bed heater, though serving perhaps for Mach numbers from 5 to 10, could not, it was clear, provide the heat required for representing the conditions encountered by reentry bodies. There was yet, however, a possibility of accomplishing this objective in a tunnel capable of operating for reasonably long periods of time. This possibility lay in the use of an electric arc to heat the air as it passed through the tunnel. The initial investigation of arc-heated jets was made in 1956 by Jeff Buck, R. W. Eglington, A. Kamiya, Merrill Nourse, and others. Later the work was continued by William Carlson and Carl Sorenson. First investigated were some arc-jet ideas  which had originated in Germany. This study, however, was just the beginning of work that was still in progress and accelerating at the end of 1957. Aside from keeping the tunnel walls and electrodes from melting, one of the problems in the design of an arc-jet facility arose from the contamination of the air by vaporized material from the electrodes. The problems in the development of a practical arc-jet tunnel were obviously great, but the need for such a facility was also great and the project was pushed with ever increasing vigor.
The use of electronic computing facilities expanded rapidly during this period. An Electrodata Datatron digital computer was purchased in 1954 and was soon being used for "on-line" data reduction for the 6-by-6 and the three Unitary Plan wind tunnels. In the on-line procedure, the electrical strain-gage balances with which wind-tunnel models were generally equipped transmitted their electrical outputs directly to the computer for the application of calibration factors and the immediate computation of aerodynamic coefficients. An earlier "off-line" procedure was to have the strain-gage readings appear on punched cards which at some convenient later date were fed into the computing machine for the computation of coefficients. With on-line computations, the results of a test were immediately available for inspection and also, if desired, an automatic plotter could be used to plot the coefficients as they came from the computer.
In 1955 an IBM 650 digital computer was leased for the specific purpose of serving the computational needs of the theoreticians in the various sections. At this time the first effort was made to train people in the various research sections to do their own computer programing.
In 1956 or thereabouts, a second Datatron computer was purchased. This computer, like the first Datatron, was used for wind-tunnel data reduction, both on-line and off-line.
Additional analog-computer elements, used in flight-simulation work, were also procured during this period. These, however, were separately located and under the control of Stanley Schmidt and Howard Matthews of the Dynamics Analysis Branch, Full Scale and Flight Research Division. The digital computers, on the other hand, remained in the Electronic Machine Computing Branch of the Theoretical and Applied Research Division. This Branch, as earlier mentioned, was headed by Dr. Mersman. The original Assistant Branch Chief, Harold Harrison, resigned in 1955 and his position was then assumed by Stewart Crandall.
1 transfer was later formalized in a letter from Arthur B. Freeman, Assistant Director for Administration, to Dr. E. W. Strong, Chancellor, University of California, Berkeley, dated Feb. 16, 1962.
2 See ch.7
3 NASA report series to be described later.
4 Designed to withstand an internal pressure of 2000 psi, this massive vessel was reported to be the heaviest single item the Southern Pacific Railroad had ever transported.