SP-4302 Adventures in Research: A History of Ames Research Center 1940-1965









[153] In deciding who should do what in aeronautical research and development after World War II, there was a great deal of interest in the definitions of research. It was commonly and loosely referred to as pure, applied, basic, scientific, fundamental, or something else, but the meanings of these terms remained obscure. Where in the broad spectrum of activity called research did the work of NACA, and Ames, really lie?

Since the object of "pure" research might reasonably be solely the satisfaction of human curiosity, the definition "pure" could not be applied to the activities of NACA. The work of the NACA laboratories would have to be regarded as "applied" research; that is the only kind of research a practical-minded Congress would have funded. It was research applied directly to solving the practical and pressing problems of military and civil aviation. It was also basic and scientific, but only to the extent that it dealt with matters of fundamental significance.

How fundamental was NACA research? If the characteristic being studied applied only to a single specimen, animate or inanimate object, its study would certainly be the least significant activity that could be called research-and even so might better be called "development" or "evaluation." But if the characteristic being studied were the charge on the electron, applying to all the electrons in the universe, or the makeup of the DNA molecule' found in the cells of every animal from cockroaches to humans, then we could truly say that such research had the highest degree of fundmentality. Between these limits lies the vast bulk of research, distributed according to the breadth of application of its results but with no sharply defined divisions. On this scale, most of NACA's work lay well below the midpoint and during the war some of it lay close to the bottom. In general the results of NACA's work applied to a single, narrow class of objects called airplanes; during the war, it sometimes applied to only one or two specific Specimens of this class. Fortunately, fundamentality is not necessarily a measure of practical value.

[154] NACA's work generally was an effective mixture of theory and experiment. Theory contributed much to the fundamentality of any research effort. It also was used to limit the extent and establish the direction of required experiment and, metaphorically speaking, to define the major limb structure of the tree of research. In general, fundamental research was a movement downward toward the root source, while applied research was a movement upward and outward toward the leaves. Often the two were profitably combined. The leaf pattern of the tree of research was filled out by experimentation, and sometimes the leaf pattern was used to deduce the position and form of a hidden limb. The limbs were of the utmost importance, of course, but it was among the leaves that the fruits usually lay.

Experimentation might be thought of as a form of observational research such as used in astronomy; but, instead of waiting for Nature to speak in her own good time and place, we ask her questions and deliberately force her to speak at a time and place of our choosing. She withholds nothing from those clever enough to ask the right questions. Our questions are asked by confronting her with a cunningly devised situation or mechanism to which she must react. Devices by means of which such confrontations are achieved are often in the nature of simulators. Members of an important class of simulators used by NACA were called wind tunnels.

In devising simulators for research, the research engineer has no equal, and the NACA laboratories, at least until this time, were laboratories of research engineers, not scientists. They were also laboratories of ingeniously designed simulators the development of which, in some cases, represented true adventures in research.




wings. By original work in TR 863 on sweep and in TR 835 on pointed low-aspect-ratio wings, R. T. Jones had provided intriguing evidence of things to come in aircraft configurations. No longer could wings be regarded merely as assemblages of largely independent airfoil sections whose action might be represented in theory by a lifting line. In the past the most significant feature of a wing was the airfoil section and perhaps no other single item had received so much attention in aeronautical research. In the new breed of aircraft, the airfoil sections tended to lose their identity and become aerodynamically and physically blended into a whole lifting surface. Now it was the wing planform, the shape of the wing as viewed from above, that was of greatest importance.

Although less important than in the past, airfoil sections were still regarded at Ames as worthy of some research attention. Before going off for a doctorate at Caltech, Milton Van Dyke, in TN 2670, summarized the work he had earlier done in the 1- by 3 1/2-foot tunnel on NACA 6-series high-speed sections; and during the early postwar period George McCullough [155] and Donald Gault were investigating, in the 7-by-10, a new type of stall that had been revealed by the thin airfoil sections required for high-speed flight The wings of airplanes had to be thin if shock-wave formation and its bad effects were to be avoided at high subsonic speeds. But wings with thin leading edges performed very poorly in generating the lift required for landing. Even at moderate angles of attack, the air tended to separate from the forward upper surface, the separation appearing first as a bubble and then spreading over the whole surface as a violent stall. McCullough and Gault studied this phenomenon and attempted to cure it by removing air through a suction slot at the point of separation.

Airfoil research was rather old hat, but what was really fun during this period was the investigation of novel wing planforms. In the early part of the period there was little theory and scant knowledge about how to design wings for transonic and supersonic airplanes; thus research people cast inhibitions aside and investigated many different wing shapes, utilizing all available facilities. Ames was in a fortunate position for such studies. It had the 12-foot, the 16-foot, and the 40- by 80-foot tunnels for large-scale subsonic tests, and the 1- by 3-foot and the 6- by 6-foot tunnels for supersonic tests. Moreover, it was prepared to use the bump and wingflow techniques for transonic studies. These facilities would allow testing through wide ranges of size, speed, and Reynolds number and would enable the Ames staff to conduct coordinated test programs in several different tunnels on promising wing configurations.

Wing Theory. In the most favored research technique, theory precedes and guides experiment, but in early 1946 supersonic and transonic wing theory had not advanced very far. Thus initial experiments at Ames were undertaken largely without benefit of theory. Nevertheless, the development of theory proceeded rapidly and there were soon notable contributions from NACA and other sources. At Ames, linearized, lifting-surface, supersonic wing theory was advanced by Max Heaslet, Harvard Lomax, and Arthur Jones, who turned out a number of papers of which TR 889 (ref. B-1) is perhaps representative. The development of transonic theory was shortly initiated by John Spreiter, who had just returned to Ames after completing work for his master's degree at Stanford. Spreiter, extending theories originally developed by Max Munk and Robert Jones, devised methods for predicting the characteristics of slender wingbody combinations at subsonic, transonic, and supersonic speeds. Later, in TR 962 (ref. B-2), he extended this work to include cruciform wing configurations such as might be used in missiles. Max Heaslet, Harvard Lomax, and John Spreiter also collaborated to produce TR 956 (ref. B-3), which was a substantial contribution to linear transonic wing theory. These efforts, though worthwhile, clearly represented only a small part of the work that needed to be done on wing theory, particularly transonic wing theory.



Harvard Lomax.

Harvard Lomax.

Dr. John R. Spreiter.

Dr. John R. Spreiter.


Wing Experiments. In wing-study programs, the experimentalists had a rich field to exploit. Of wing variables, they could play with sweep (both forward and backward), aspect ratio, taper ratio, camber, twist, and even airfoil section. Some of the sections investigated were sharp edged, flat sided, and in the form of a very elongated diamond or double wedge. A major objective of such studies was to delay as long as possible the drag rise and other adverse effects arising from the formation of shock waves and then to minimize the severity of these effects where, at the higher speeds, they could not be completely avoided. It appeared that a swept wing of otherwise conventional configuration might serve in the high subsonic and low supersonic speed ranges but for higher supersonic speeds a thin low-aspect-ratio wing would be desirable. Sweep tended to limit aspect ratio and, when it became extreme, there were structural and other practical reasons for filling in the space between the wing and the fuselage. In this way the triangular wing shaped like the Greek letter delta was developed. The delta wing provided sweep, low aspect ratio, structural stiffness, and internal space for carrying fuel. One of its early applications was in the Consolidated XF92A, which incorporated a triangular wing and triangular vertical tail.

At Mach numbers of 1.5 and higher, a simple, thin, low-aspect-ratio straight wing seemed to give good results, and such a wing was incorporated in the Douglas X-3, the design of which was commenced in 1946 or 1947 The X-3, conceived by Francis Clauser, Bailey Oswald, Schuyler Kleinhans,....



Wing planforms.


Drag of low-aspect-ratio wings.

Drag of low-aspect-ratio wings.


....and others at Douglas Santa Monica, represented perhaps the first attempt made in this country to design and build a truly supersonic airplane capable of sustained flight and unassisted takeoffs. Only later, after the development of the airplane had been long delayed by lack of funding, did it become one of the research airplane series. Its daring conception was a landmark in aviation although not adequately recognized as such.

While investigating many wing configurations, Ames devoted more of its attention to triangular wings than to wings of any other planform. Triangular wings incorporating a wide variety of aspect ratios, thicknesses, airfoil sections, and shapes were tested in the Ames facilities. The shapes varied from the delta configuration with tips aft, to the reversed delta with tips forward, and to other configurations having tips at intermediate locations.

Another fairly extensive program of tests was run on a rather special Configuration incorporating long tapered wings swept backward 63°. The great interest shown in this particular design stemmed from Bob Jones' [158] report, TN 1350 (ref. B-4), which indicated the possibility of designing an airplane that, at moderately supersonic speeds, would fly with almost subsonic efficiency. This concept was rather intriguing, as supersonic flight had been thought of more in terms of brute force than of efficiency. The achievement of high efficiency (measured in terms of L/D or lift-drag ratio) would, according to Jones, require long, slender, uniformly loaded wings swept sufficiently to fall within the conical shock pattern (Mach cone). Thus the leading edges of the wings would lie in a region of subsonic flow and perhaps could be expected to develop the same leading-edge suction that increased the lifting efficiency of subsonic wings. Accordingly, test models were built of a wing-body combination designed to operate at a Mach number of 1.53 and at L/D, it was hoped, of 10 or more. The wings had an aspect ratio of 3.5, a taper ratio of 0.25, were swept back 63°, and were twisted and cambered to give a uniform loading along the span. These models, and models with slight variations, were tested in several wind tunnels at Ames; but the elusive leading-edge suction upon which high efficiency was dependent was never fully realized. A representative report covering work performed in this program is RM AJ324 by Charles Hall and John Heitmeyer.

As earlier mentioned, most of the Ames wind-tunnel groups contributed to the general study of wing planforms. The 12-foot tunnel was able to test good-sized models at either high subsonic speed or high Reynolds number. George Edwards, Jack Stephenson, and Bruce Tinling investigated and....


Twisted and cambered 63 degrees swept wing.

Twisted and cambered 63° swept wing.


[159] ....reported on triangular wings. Ben Johnson concerned himself with straight low-aspect-ratio wing, while Bruce Tinling, Robert Reynolds, Donald Smith, Lloyd Jones, and Fred Demele investigated swept wings, including R. T. Jones' 63° wing. In the course of large-scale tests conducted in the 40-by-80, Gerald McCormack, Victor Stevens, and Woodrow Cook investigated wings swept forward and backward, while Lynn Hunton and Joseph Dew studied the effect of camber and twist on the loading and stalling characteristics of a 45° swept wing.

The Flight Research Section at Ames also contributed to the Laboratory's wing-research program. Its contribution was made through the use of a P-51 airplane in an application of the wing-flow technique. This effort, carried out by George Rathert and Carl Hanson, included tests of a straight, low-aspect-ratio wing of the same configuration as a larger model tested in the 12-foot tunnel. It also included tests of a delta-wing model.

When the 1-by-3 got under way late in 1945, there was a feeling, according to Dean Chapman, that operational supersonic aircraft were still a long way off. The transonic range, with its unpleasant compressibility effects, then appeared to be a rather formidable barrier to supersonic speeds. Nevertheless, immediate steps were taken to lay out a program of tests on a number of wing planforms which were thought suitable for supersonic flight. The selection of the planforms to be tested was rather arbitrary since little theory had then been developed. Walter Vincenti, with the help of Elliott Katzen and others, took the lead in carrying out this program. Great pains were taken to coordinate the test results with existing theory and to check theory against experiment. Two important reports resulted from this work: RM A7I10 (ref. B-5) by Walter Vincenti, Jack Nielsen, and Fred Matteson, which presented data on both normal and reversed delta wings, and TR 1033 (ref. B-6) by Vincenti, in which a critical comparison was made of theory with experimental data obtained on a variety of straight-wing and swept-wing models. The tests were all carried out at a Mach number of 1.53 using a fixed-throat nozzle in lieu of the variable nozzle which was still under development.

In the 6- by 6-foot tunnel, Hall and Heitmeyer, as earlier noted, investigated the 63° wing, while Charles Frick and R. S. Chubb made an important analytical study of the longitudinal stability of elastic sweptwings at Supersonic speeds. This study, reported in TR 965 (ref. B-7), pointed to the very important role played by structural elasticity in the aerodynamic behavior of large sweptwing airplanes. Wing thinness, required for high-speed flight, and sweep each contributed to wing deflections under load and thus to the static stability and dynamic response of large sweptwing airplanes such as the B-47. Aeroelastic effects on airplane performance were not new, of course, but never before had they assumed such importance as in this new regime of swept wings.

[160] Controls. The 16-foot tunnel during this period undertook quite a lot of work on controls for sweptwing and other high-speed aircraft. Lee Boddy collaborated with Walter Williams of the NACA group at Edwards in preparing a summary and analysis of NACA's work on dive-recovery flaps. This study was published as RM A709. In the same period John Axelson prepared a summary and analysis of wind-tunnel data on the lift and hinge-moment characteristics of control surfaces at Mach numbers up to 0.9, while Walter Krumm and Joseph Cleary produced a series of reports on the lateral control of straight- and swept-wing airplanes. The problem of devising controls for swept wings was particularly difficult owing to the tendency of the boundary layer on such wings to flow spanwise toward the tips, thus causing early stall in the region of the ailerons. To prevent such lateral flow, boundary-layer barriers, called "fences," were often installed at certain points on the upper surface of the wing.

Another very useful bit of control-system work done in the 16-foot tunnel during this period concerned a very serious stability problem that the Douglas Santa Monica people had encountered in early flights of their new C-74 transport. NACA's help was requested and a large segment of the airplane's horizontal tail surface was installed in the 16-foot tunnel. Although it first appeared that the blockage produced by the huge test body might negate NACA's test efforts, the first run in the tunnel revealed the cause of the problem. At high speed, aerodynamic forces caused ballooning of the fabric covering the elevator, thus greatly affecting the hinge moments and controllability of the airplane. To reduce the weight and to facilitate the balancing of movable control surfaces, it had been common practice to cover them with fabric rather than metal. The experience with the C-74 proved that the day was over when this practice could be tolerated; indeed, the day had really been over for some time. The same control problem had earlier been experienced by other airplanes.

Bodies. Although the Laboratory during this period was preoccupied with wing research, there were also several notable investigations relating to bodies (fuselages). Indeed, among the significant early programs run in the 1- by 3-foot tunnel were a number that were concerned with the effects of viscosity on the supersonic flow over bodies of revolution. These studies were made by Dean Chapman, Edward Perkins, and Harvey Allen. The 1-by-3 was particularly useful for investigations of this kind because of the ease with which its operating pressure could be varied. This favorable operating characteristics greatly facilitated the study of Reynolds-number effects.

Perhaps the first report written at Ames on supersonic tests was a paper by Dean Chapman and Ed Perkins on the effects of viscosity on the drag and base pressures of bodies of revolution at a Mach number of 1.5. This paper, which later was published as TR 1036 (ref. B-8), gave considerable atten-....



Vapor-screen technique, showing position of vortices.

Vapor-screen technique, showing position of vortices.


...-tion to the flow in the mixing region at the truncated base of a body. The study so aroused Chapman's interest in base-flow phenomena that, when he left shortly thereafter to complete his doctoral work at Caltech, he adopted the base-pressure problem as his thesis subject. Later, on his return from Caltech, his thesis work was included in TR 1051 (ref. B-9), "An Analysis of Base Pressure at Supersonic Velocities and Comparison With Experiment."

While Chapman was at Caltech, Harvey Allen and Ed Perkins made a combined theoretical-experimental study of the flow around an inclined body of revolution. This study, reported in TR 1048 (ref. B-10), provided an approximate method for calculating the force and moment characteristics of inclined bodies and revealed the existence of a pair of vortices shed from the body on its lee side. In tests run in the 1- by 3-foot tunnel, the vortices were made visible by a vapor-screen technique devised by Allen.

In the vapor-screen technique, a small amount of water is introduced into the tunnel; the water vaporizes and condenses into a fog in the test section. A plane of intense light is then passed through the schlieren windows and thence transversely through the test section in the region of the model. The vortices act like a centrifuge on the water vapor and their cores, as they pass through the light screen, leave dark spots in an otherwise glowing sheet of diffused light. Other features of the flow pattern are also revealed. In this and other instances, the study of vortex shedding from bodies and wings was considerably facilitated through the use of the vapor-screen technique.

Meanwhile in the 16-foot tunnel, Lee Boddy and Charles Morrill were investigating the possibility of contouring a fuselage in such a way as to minimize the flow interference at its juncture with a swept wing. This work, which was somewhat prophetic, gave recognition to an incompatibility at the wing-fuselage juncture between the three-dimensional flow pattern over a Swept wing and the essentially two-dimensional flow pattern over the fuselage.


[162] INLETS


With the advent of jet engines, the problems of air inlets and internal aerodynamics became of major importance and of great influence in airplane design. The problem of where to locate the engines and the inlets had to be solved; then the detailed design of the inlets and internal flow systems seemed a subject for vital and almost endless research. There appeared in this period to be three promising locations for jet air inlets: in the nose of the fuselage, in the sides of the fuselage, and in the wing roots. Inlets in each of these locations were investigated at Ames in a program thoroughly coordinated with similar work in the Langley and the Lewis Laboratories.

The Ames inlet investigations were conducted in a number of facilities including the 7- by 10-foot tunnel; an 8- by 36-inch flow channel constructed at the 7-by-10 especially for inlet studies; the 16-foot, the 40- by 80-foot, and the 1- by 3-foot tunnels; and the 8- by 8-inch tunnel, which originally was built to check out the sliding-block nozzle but which had since proved very useful for research work. A rather substantial effort was made by the Laboratory in perfecting the submerged inlet developed by Mossman and others in the 7-by-10. This work was carried out at reduced scale in the 7-by-10 flow channel, at full scale in the 40-by-80, and at transonic speeds by bump tests in the 16-foot tunnel. In the 1-by-3, a group headed by Wallace Davis concentrated on side inlets of the scoop variety; and in the 8- by 8-inch tunnel, John Lundell and others investigated supersonic nose inlets, particularly those applying to ramjet-powered missiles.

Among the more significant reports on submerged inlets were RM A7130 by Emmet Mossman and Lauros Randall and, from the 16-foot tunnel, RM A8B16 by Charles Hall and Dorn Barclay. In the 40-by-80, Norman Martin and Curt Holzhauser investigated twin side scoops which fed a common engine inlet duct, and in TN 2049 they analyzed the instability that was found to occur in such systems at low flow rates.




Ames engineers were concerned with the dynamic as well as the static stability of new airplane configurations, particularly those required to operate in the transonic and supersonic speed ranges. In the 6- by 6-foot tunnel, triangular wing models were spring mounted, free to pitch, and their damping characteristics measured. This work was reported in RM A50J26 by Murray Tobak, Dave Reese, and Ben Beam.

The flutter of ailerons, excited by oscillating shock waves, had by this time become a fairly common and troublesome phenomenon. Inasmuch as existing knowledge of the subject was rather superficial, Albert Erickson and Robert Robinson of the 16-foot tunnel undertook to learn more about the aerodynamic forces involved. Toward this end they devised special instrumentation and techniques with which they were able to measure instantaneous [163] pressure distributions over a representative section of the aileron while it was in the very act of fluttering. To do this, they enlisted the aid of Jim Kyle, a fellow member of the 16-foot-tunnel staff, who designed a tiny, fast-acting pressure cell that could be mounted flush with the surface of the aileron and wing. This interesting instrumentation and technique was described by Erickson and Robinson in RM A8H03.

As part of a general wing-study program during this period, the 40- by 80-foot tunnel section carried out some very important work having to do with the development of an analytical method for determining the distribution of airloads on swept wings at subsonic speeds. This problem had been solved for straight wings but, in the case of swept wings, it was much more difficult. John DeYoung was the major contributor to this effort and was helped by Charles Harper and Victor Stevens. Key reports issuing from this work include TR 921 (ref. B-11) by DeYoung and Harper and TR 1056 by DeYoung alone. While this work was going on in the 40-by-80, Doris Cohen (Mrs. R. T. Jones) of the 1- by 3-foot tunnel section was engaged in making theoretical determinations of the air loads on swept wings at supersonic speeds. Her work is covered in TR 1050 and other papers.




Deicing. In the early postwar period the work of the Ames Flight Engineering and Flight Research groups was mostly a continuation of what they had been doing during the war. Icing research continued for a number of years under the able guidance of Alun Jones but became largely concerned with perfecting analytical techniques, obtaining more complete meteorological data on icing conditions, and developing deicing techniques for such airplane components as windshields and propellers. The study of carburetor icing had been taken over by the NACA Lewis Flight Propulsion Laboratory. Major contributors to the icing studies at Ames during this period were Alun Jones, Carr Neel, Norman Bergrun, James Selna, and George Holdaway of the Laboratory staff, and William Lewis of the U.S. Weather Bureau. The fine work done by Lewis throughout the icing research program was recognized in 1949 when he was given the Robert M. Losey Award of the Institute of the Aeronautical Sciences. As the period ended in 1949, the Flight Engineering group was diverting its efforts from deicing to the development of the drop-test method of transonic flight research. This development has earlier been mentioned.

Flying Qualities. Although the wind-tunnel people at Ames were Working on some very advanced airplane configurations, none of these configurations had yet appeared in airplanes which the Ames Flight Research group were testing. It was not until 1949 that the F-86, the first sweptwing airplane, came to Ames for flight tests. Until then, the fastest airplane in the Flight Research stable was the straight-wing P-80A which in a dive, as [164] Clousing found out, could reach a Mach number of about 0.88. Rather extensive flying-quality tests were run on the P-80A, and these were reported by Seth Anderson, Frank Christofferson, and Lawrence Clousing in RM A7G01. Larry Clousing's outstanding work in flight research was recognized in 1947 when the Octave Chanute Award was conferred on him by the Institute of the Aeronautical Sciences.

Other flight research at Ames during this period represented a continuation of earlier efforts to confirm in flight the stability and control of airplanes as predicted by wind-tunnel tests. There was also an application of the wing-flow technique to transonic studies both of straight wings, such as used on the X-3, and delta wings.

Loads. Also carried over from the war years was the program on wing and tail loads in which Larry Clousing, Melvin Sadoff, William Turner, and others were playing such an active part. More recently the Ames Flight Research Section had begun an investigation of buffeting-an unpleasant, if not dangerous, phenomenon encountered by high-speed airplanes in dive pullouts. The buffeting, occurring at high subsonic speeds, appeared to arise from oscillating shock waves which caused cyclic separation of the flow on the upper surface of the wing. The conditions under which the buffeting occurred were investigated on a number of different airplanes and it was while testing one of these airplanes, the P-51, that George Cooper, the pilot, thought he could actually see the shock wave on the wing. Further investigation showed that what he had seen was indeed the shock wave or, more precisely, a refraction pattern caused by the sun's rays shining through the shock wave. The shock wave was then seen on the wings of two other airplanes and the conditions under which it would appear-the angle of the airplane with respect to the sun, etc.-were determined. Photographs were taken of the shock waves and it was noted that the shock wave would move backward, toward the trailing edge, as Mach number increased and forward as lift coefficient increased. It was also noticed that, at some value of Mach number and lift coefficient, the shock would dance back and forth through an amplitude of 2 inches or more. Moreover, buffeting appeared just when the dancing began. As the technique of visually observing shock waves was thought to be useful, George Cooper and George Rathert wrote a report on the project which was published as RM A8C25.

Variable-Stability Airplane. By far the most important contribution of the Ames Flight Research group during this period was the development of the variable-stability airplane. In this project, William Kauffman played the leading part while G. Allan Smith and others played important supporting roles. Apparently the idea of a variable-stability airplane was not new (it had been mentioned in German literature), but it was one of those things the practical accomplishment of which requires greater genius than the original conception.



Visible shock waves in flight

Visible shock waves in flight (above) and William M. Kauffman (below).

William M. Kauffman


Aside from its application to the wing-flow method, the airplane had generally not been used in the past as a simulator. In flight it represented only itself; it was indeed the ultimate yardstick against which wind-tunnel model tests were measured. And the airplane, of course, could provide information on flight dynamics, on the interrelationships between pilot and airplane, that were quite beyond the powers of any wind tunnel, or any other simulator, to produce. But was this statement precisely true and could not the airplane itself be made a simulator, to simulate the dynamic behavior and pilot-machine relations of other airplane configurations? The concept was most interesting. Such a simulator would be a wonderful tool for studying the flying qualities of the radically different airplane configurations that shortly would be in use. But was the development of such a simulator a practical possibility? It certainly was, said Kauffman, and let's get at it. Immediately enlisted in the project was Dr. G. Allan Smith, expert on servomechanisms, from the Instrument Research Section. The conservative propeller-driven Grumman F6F would be used in this first attempt to produce a variable-stability flying simulator.

Of greatest interest at the moment were the lateral-directional dynamics of airplanes; thus the study of this area of airplane performance was Selected as the first application of the variable-stability airplane. But to simplify the variable-stability airplane development problem, only one aspect of lateral-directional performance would be investigated-the effect of changing the wing dihedral. The physical alteration of the dihedral in any airplane represents a major structural modification, but the effect of changing [166] dihedral could be obtained by installing a powered servomechanism in the test airplane that would deflect the ailerons in proportion to the angle of yaw. However, it would be necessary for the device to accomplish this objective without moving the control stick or changing the stick force required for the pilot to operate the ailerons. In others words, the pilot would not sense that the device was operating except for the fact that the dihedral of the airplane would appear to have been changed. The design of the device to accomplish this objective was very tricky and had to be worked out with great care if the plane and pilot were not to be endangered. But the job was done and done well. By the adjustment of the mechanism in flight, any effective dihedral from -18° to +28° could be obtained.

The variable-stability airplane was flown by five different pilots under conditions simulating landing approach, cruising, and high speed. Under each of these conditions, the effective dihedral was changed through a wide range and the opinion of each pilot as to the quality of the airplane performance at each dihedral setting was noted. In this way, the optimum dihedral and the tolerable range of dihedral were ascertained for each condition of flight. Special flight instrumentation provided useful quantitative data, of course, but in the end it was the pilot's opinion of the flying qualities of the airplane that counted most. Clearly, pilot opinion would become increasingly important as the new class of airplane, having radically different inertia and control characteristics, came into use. For such airplanes, past experience and existing data were largely inapplicable. In any case, the very significant story of the conversion of the F6F and the results of the first tests were reported in TR 948 (ref. B-12) by Kauffman, Liddell, Smith, and Van Dyke. Rudolph Van Dyke was one of the Ames pilots who flew the airplane.

The results of this first use of a variable-stability airplane by NACA at Ames were very promising and extensions of the idea quickly came to mind. To obtain a better simulation of the whole lateral-directional stability syndrome, it would be necessary to drive the rudder, as well as the ailerons, in accordance with yaw angle and also, perhaps, to introduce roll rate and yaw rate as inputs in addition to yaw angle. Steps to accomplish these changes were soon taken.

Sonic Boom. The F-86 was the first operational airplane capable, in a dive, of reaching supersonic speeds. The XS-1 (now called the X-1), carried aloft by a mother airplane, had of course achieved supersonic speed in level flight at high altitudes over the desert. Ames received one of the first F-86's for flight tests in 1949 and shortly was running tests involving prolonged dives at very high speed. Rudolph Van Dyke and George Cooper were the pilots. Soon after the program got under way, the local newspapers began reporting mysterious explosions the source of which could never be located. One day such an explosion shook the plates off the plate rail in the International.....



Forced landing of a Douglas XSB2D-1 airplane in a prune orchard by George Cooper and Welko Gasich.

Forced landing of a Douglas XSB2D-1 airplane in a prune orchard by George Cooper and Welko Gasich.


....Kitchen near Niles, Calif. A sheriff's posse was sent out to investigate the cause but had no luck. Another explosion occurring in the region of the Calaveras Reservoir seemed to be focused by a box canyon and reportedly caused some slight damage to a house at the head of the canyon.

Newspaper headlines reporting explosions became bigger and blacker. McAvoy, walking into the pilots' office with a newspaper in his hand one morning, jokingly asked Van Dyke and Cooper, "What are you fellows doing to cause so much noise?" Mac's remark was a bit of purely innocent humor but it caused something to click in Van Dyke's mind. Good Lord! Could the explosions have something to do with their dives in the F-86? Smitty DeFrance had similar suspicions and, as he sat in his office one day, he heard a boom or explosion such as had been reported in the papers. He noted the time and called the Flight Research Section to see if the F-86 was up. Yes, it was, and yes, it had made a scheduled dive at the same minute that the boom had occurred. The suspicions of Smitty and Rudy Van Dyke were confirmed The explosions had been caused by the F-86 diving beyond the speed of sound. Ames, as a matter of policy, did not publicize the matter, but it was not many months before the sonic boom was observed elsewhere and associated with the supersonic flight of airplanes. Soon the boom became commonplace; but Ames, it is believed, was first to encounter and recognize the phenomenon

Hazards In dive testing the F-86, Ames pilots Cooper and Van Dyke were probably flying through the transonic range more frequently than anyone else in the country. Other hazards were encountered with slower air-....



Test pilot George Cooper.

Test pilot George Cooper.


....-planes-for example, the incident that occurred one day while George Cooper and Welko Gasich were up in a Douglas SB2D carrying out some mission over the Los Gatos countryside. The engine, demonstrating a habit of this particular model, backfired and started a fire in the induction system. George could not get back to the field, so he set the plane down between rows of trees in a prune orchard. The plane's wings clipped the treetops off neatly in a descending path; but, as the wings got down to the trunks, the going got rough. The airplane was a wreck by the time it came to rest; George and Welko walked away from it badly shaken but not seriously harmed.

Another time George was up in a P-47 investigating the effect of reversing the propeller pitch on the stability and control of the airplane in a dive.

His engineering pals on the ground assured him that "all you have to do if you get in trouble is to push this button here and the prop will snap back to normal pitch." Well, in one dive, though he was not in any trouble, he wanted to discontinue the test, so he pushed the button. Nothing happened. The prop did not move a degree and continued its devilish braking effect. George was then faced with the miserable task of making a forced landing with the propeller reversed. He carefully made his approach and, when he was only a couple of hundred feet from the runway, the propeller suddenly snapped into normal pitch. He slammed the throttle open and with a roar took off over the treetops to make another, normal, and successful landing.

Incidents of the kind mentioned were all in the day's work for Ames pilots. Unhappily they did not all turn out so well. Luck ran out for Ryland D. Carter, Ames test pilot, as he was flying a P-51H airplane on May 17, 1948. The wing of the airplane, incidentally, was equipped with a glove and otherwise adapted for wing-flow tests. For reasons unknown, the plane came apart in the air, its fragments scattering widely over farmland near Newark, [169] Calif. Although Carter was thrown clear, he was unable to activate his parachute and was killed.




Boundary-layer heating resulting from skin friction and air compression had been of rather small concern to aerodynamicists, but now became a matter of some importance when supersonic flight speeds were being considered. Aerodynamicists at Ames, and elsewhere, were interested in the effects of aerodynamic heating on the structure and contents of flight vehicles as well as in such related problems as the heat transfer to the body from laminar and turbulent boundary layers, the effect of heat transfer to, or from, the body on boundary-layer transition, and the effect of boundary-layer heating on skin friction.

At Ames there had been some early work on heating; this was reported by Allen and Nitzberg in TN 1255 and by Tendeland and Schlaff in TN 1675. In 1947, Richard Scherrer made a theoretical study, reported in TR 917, of the effects of aerodynamic heating on bodies of revolution at supersonic speeds and, in 1949, Dean Chapman and Morris Rubesin presented a paper to the IAS (ref. B-13) on temperature and velocity profiles in the compressible laminar boundary layer with arbitrary distributions of surface temperature. At about the same time, Richard Scherrer and William Wimbrow ran some tests in the 1- by 3-foot tunnel on heated and unheated cones. This work, which compared experimental results with theory, was summarized by Scherrer in TR 1055 (ref. B-14). In general it appeared that experiment confirmed the earlier theories, including one which predicted that a laminar boundary layer would be stabilized if the body were cooler than the surrounding airflow and destabilized if the reverse were true. This factor would have to be considered in future wind-tunnel model tests.

Jack Stalder and Glen Goodwin, as earlier noted, had built a very special, if low-cost, tunnel to operate at very low air densities, such as those to be found at altitudes of perhaps 50 to 70 miles, where future missiles might conceivably fly. This realm of flight, called the "free molecule" range, is a region wherein the air molecules are far apart and the average distance they travel before bumping into another molecule is considerably greater than the length of the missile or the model in the wind tunnel. Jack admitted that simulation of a somewhat lower altitude range might have been a little more practical; but the free-molecule range had the distinct advantage of being more amenable to theoretical analysis than the lower ranges. Before the low-density tunnel was completed, Jack and David Jukoff had made an analysis of heat transfer to bodies traveling at high speed in the upper atmosphere. This study, published as TR 944 (ref. B-15), dealt with air [170] molecular energy transport to a body and developed a general method for calculating surface temperatures in steady high-speed flight in a rarefied atmosphere.

When the low-density tunnel was completed, and the difficult problems of devising low-pressure instrumentation were solved, tunnel calibration tests began. It was found that the boundary layer on the walls expanded under the influence of the low pressure and nearly filled the throat. Only a 1/2-inch-diameter core of untrammeled air at the center remained. This condition considerably limited the test possibilities of the tunnel. Jack and Glen decided that their first test, a test with which theory could be checked, would be of the aerodynamic heat transferred to a cylinder mounted transversely across the flow. The cylinder was actually a wire; indeed, it was an iron-constantan thermocouple with the butt-welded junction located in the core of the tunnel airstream.

The results of the tests in the new tunnel were classic. They confirmed the scarcely believable indications of theory, that the temperature of the wire in the low-density airstream should be higher than the at-rest temperature of the air in the tunnel. At pressures and densities found in any ordinary tunnel, the maximum temperature the wire could be expected to reach would be the at-rest, or stagnation, temperature of the tunnel air; but, at low densities, the wire temperature was as much as 65° F higher than the stagnation temperature of the tunnel when nitrogen, a diatomic gas, was used in the tunnel and as much as 147° F higher when helium, a monatomic gas, was used. The drag on the wire, though exceedingly small, was also measured and found to confirm theory.

The results of these first tests alone justified the money and time spent on the tunnel. They are contained in TR 1032 (ref. B-16) by Jack Stalder, Glen Goodwin, and Marcus Creager. The report not only develops the theory of the wire heating but also gives the confirming experimental results. In addition, it provides a good description of the tunnel and its auxiliary equipment.

The new 6-inch heat-transfer tunnel also was put into operation in 1949. The first test report to come from it was TN 2077 by Jack Stalder, Morrie Rubesin, and Thor Tendeland. The subject of this report, which included a description of the tunnel, was the temperature-recovery factors on a flat plate in a supersonic airflow.