SP-4305 ENGINEER IN CHARGE

 

4

With a View to Practical Solutions

 

[97] Though Munk resigned from the NACA in early 1927, Langley's systematic experimental program to develop improved airfoils in the VDT continued unabated. In 1933 the Committee published "Characteristics of 78 Related Airfoil Sections from Tests in the Variable-Density Tunnel." This report introduced what was to be the VDT's principal achievement as an aeronautical research tool: a second and more significant series of airfoils, the NACA 4-digit series.

 

The Mind's Eye

 

In the course of developing the second airfoil series in the late 1920s and early 1930s, the VDT team devised a numerical code-patterned after that used to identify the composition of steel alloys-by which to describe the physical shapes. Like all other aerodynamical laboratories, Langley had until then designated airfoils simply by numbering them in the sequence in which they had been tested (M-1, M-2, M-3, and so on). In the new system, however, four numbers would indicate the airfoil section's critical geometrical properties. The first integer represented the maximum mean camber in percent of the chord; the second integer represented the position of the maximum mean camber in tenths of the chord from the leading edge; and the last two integers represented the maximum thickness in percent of the chord. Thus, airfoil "N.A.C.A. 2415" was a wing section having 2 percent camber at 0.4 of the chord from the leading edge, with thickness 15 percent of the chord. Zeroes were used for the first two integers when the section was symmetrical, as in the case of N.A.C.A. 0015. The laboratory expanded the code to five and then six digits for subsequent airfoil series, and indicated modifications like changes of the leading-edge radius or the position of maximum thickness by adding a suffix consisting of a dash and...

 


[
98]

The NACA's four-digit airfoil family, 1929.

By August 1929, tests in the Variable-Density Tunnel had derived the family of airfoils N.A.C.A. 0006 through N.A.C.A. 6721, shown here in cross section.

 

....two more digits, as with N.A.C.A. 23012-64, an outstanding section in the popular 230-series. 1

This code did not signify much to the man on the street, but to aeronautical engineers it suggested everything important about an airfoil. The NACA's 1933 report on 78 related airfoils, which formally introduced the numbering system, became a classic, a designer's bible. From the mid-1930s on, one could say, for instance, "N.A.C.A. 2415," and an airfoil complete with a camber line, position of maximum thickness, and special nose features would appear in any aerodynamicist's mind's eye. Serving to remind as much as to instruct, the NACA's airfoil report complemented the coded information with graphic illustrations of two independent sets of curves. These curves communicated knowledge basic to an engineer's understanding of the relationships among an airfoil's variables.2 Graphic representation of airfoil data-the outline of the physical shape reinforced by performance curves and the digital code gave aeronautical engineers ready access to the wide range of parametric data necessary to their work. The NACA digest gave them a "whole range of wings from which to choose, the way one might select home furnishings or automobile accessories from a catalog."3 From that catalog, the American aircraft industry picked NACA airfoils that became the wings for some of the best aircraft of their era [99] including the DC-3 transport and the B-17 Flying Fortress, as well as a number of postwar general aviation aircraft.

 

Tunnel Turbulence

 

In 1929 C. G. Grey, British engineer and editor of The Aero plane, attended ceremonies at Kitty Hawk celebrating the 25th anniversary of the Wright brothers' flight. During his stay in America, Grey visited the NACA laboratory in nearby Hampton. Upon his return home, he upset his British colleagues by expressing the opinion that "the only people so far who have been able to get something like accurate results from wind tunnel experiments are the workers at the experimental station at Langley Field." 4

The NACA staff had made a reputation by building and making good use of the VDT and a few other unprecedented facilities. By that time, the Propeller Research Tunnel (PRT) had made its initial contribution to the development of a low-drag engine cowling (see next chapter), research work had begun in an 11-inch high-speed tunnel which used exhaust air from the VDT (see chapter 9), and a giant 30 x 60-foot full-scale tunnel was under construction. Above all, however, it was the VDT, representing the Committee's first bold step in the direction of novel research equipment, which won the NACA its international reputation as a technologically outstanding research organization.

Ironically, though, as useful as it was, the VDT was far from the total aerodynamic triumph trumpeted in the NACA brochures: the compressed air machine suffered from intense airstreams turbulence (small-scale eddies and cross-current swirls) resulting from its small-contraction-ratio, double return design and relatively inexpensive synchronous drive motor, which followed small but rapid frequency fluctuations. These motor fluctuations made airspeed control a serious concern for tunnel operators.

Langley chief physicist Fred Norton, who had so many problems with Munk over the design and construction of the tunnel, had in fact identified the VDT's basic defect as early as April 1921. In a letter to NACA headquarters, Norton had asserted "the probability that the steadiness of flow in the compressed-air tunnel because of the small room required [to turn the airstreams] would be inferior to that in the usual type tunnel, thus considerably decreasing the accuracy of the test." 5

In spite of these shortcomings, VDT researchers were extremely proud of their facility because they knew that no one in the world had a similar instrument for penetrating the vagaries of scale effects, meaning that everyone else was getting data even less accurate than they were. By the late 1920s, however, the VDT was fast losing its edge over other wind....

 


[
100]

Plan to correct VDT defects, 1928.

This diagram, based on an LMAL drawing from 1928, illustrates the lab's plan for correcting the turbulent airflow that had plagued the original VDT. Notice in particular the change from an open-throat to a closed-throat test section.

 

 

....tunnels. Enhancing the tunnel by rectifying its limitations became critically important to the NACA staff.

In August 1927 a broken light bulb sparked a fire and explosion in the VDT and gave Langley the opportunity to tear the compressed-air machine apart and rebuild it with a closed-throat test section and a new direct-current, variable-speed drive system. But after some five years of sporadic reconstruction, the head of the VDT section, Eastman N. Jacobs, informed the engineer-in-charge that the tunnel's basic design precluded the "possibility of obtaining the steady, constant, and uniform airstreams sought in modern wind tunnels."6

NACA Langley's growing recognition of the seriousness of turbulence in the VDT was only one good reason to seek funds in 1928 for the construction of a full-scale wind tunnel. Though the Committee continued to believe in the VDT as "a satisfactory means for testing the component parts of an airplane" and, in particular, "for conducting fundamental research on airfoil sections," it also wanted a state-of-the-art facility large enough [101] to permit the testing of actual pursuit aircraft complete with operating engines and slipstream effects. Fortunately for the NACA, Congress decided to appropriate the money for the construction of the Langley Full-Scale Tunnel (FST) in February 1929, just months before the Wall Street crash. Contracting for materials and labor at Depression prices, the laboratory was able by May 1931 to complete what was then the world's largest wind tunnel at a cost of just over one million dollars.7

 

FST vs. VDT Debate

 

The need to investigate the degree of dynamic similarity between the performance of scale models in small tunnels and the performance of airplanes and components at full scale led the FST section into spirited competition with the VDT group. Preliminary tests in the 30-x 60-foot facility convinced the FST staff that turbulence in the new machine was so "unusually low" --certainly much lower than in the VDT-that its effects could "be neglected in applying the data to design." 8 Then the results of an FST study of the characteristics of several large airfoils of various designs, including some from the NACA 4-digit family, indicated an increase in drag caused by differences in section thickness, a key design parameter, at a rate much less than that predicted by VDT tests. This was a discrepancy that directly affected the choice of wing thickness for the inner sections of airplane wings.9 As more and more tests in the FST showed good agreement with results obtained in flight, some of the prouder and less circumspect proponents of the FST even went so far as to contend that results from the VDT bore little relation to what really happened in flight and that correct airfoil data could only be obtained from tests on full-scale wings in the FST. VDT defenders, though fully aware by this time of their facility's inherent defects, answered the charges of their peers by asserting that their machine was still the NACA's best cheap means of obtaining a wide range of comparative data on a multitude of related airfoils. FST test specifications called for aircraft and aircraft models that were simply too cumbersome and expensive, they argued, to permit the kind of systematic research programs that had been accomplished in the VDT.10

The FST vs. VDT debate continued into the mid-1930s, stimulating members of all LMAL wind tunnel teams to think about the factor of scale and the corrupting effects of turbulence on aerodynamic measurement. In particular, however, the debate seems to have sparked the ingenuity of the VDT team itself, whose work was most in question. Many of these researchers began to look more carefully at the flow phenomena, especially in the boundary layer, that might be the source of the consistent errors in [102] their results. (The boundary layer is the thin stratum of air very close to the surface of a moving airfoil in which the impact pressure-that is, the reaction of the atmosphere to the moving airfoil-is reduced because of the air's viscosity. In this layer, which is separated from the contour of the airfoil by only a few thousandths of an inch, the air particles change from a smooth laminar flow near the leading edge to a more or less turbulent flow toward the rear of the airfoil. See von Karman, Aerodynamics, pp. 86-91.) To visualize the nature of the airflow around airfoils and other objects, they constructed-next to the other equipment in the VDT building-a small low-turbulence smoke tunnel. Photographs of the smoke flowing around test models facilitated study of the conditions of the boundary layer as they changed from low-friction laminar flow to high-friction turbulent flow. LMAL engineers accelerated their pursuit of a means of removing air from the boundary layer through slots or holes in the wing surface-an effort which dated back to 1926, and which was intended to decrease drag and increase lift by postponing transition from laminar to turbulent flow. Work in the smoke tunnel eventually led NACA aerodynamicists to the conclusion that two of the critical factors causing transition, and thus high skin-friction drag, were surface roughness (the rivet heads, corrugations, and surface discontinuities then common in manufactured airplane wings) and pressure distribution on the wing surface.11

Eastman Jacobs, head of the VDT section, answered the FST challenge to the integrity of VDT results by introducing the concept of "effective Reynolds number." In essence, this was Jacobs's stopgap effort to reproduce the aerodynamic effect that would be obtained in the VDT if it had zero turbulence:

 

In a wind tunnel having turbulence, the flow that is observed at a given Reynolds number... corresponds to the flow that would be observed in a turbulence-free stream at a higher value of the Reynolds number. The observed coefficients and scale effects likewise correspond more nearly to a higher value of the Reynolds number in free air than to the actual test Reynolds number in the free stream. It is then advisable to refer to this higher value of the Reynolds number at which corresponding flows would be observed in free air as the effective Reynolds number of the test and to make comparisons and apply the tunnel data at that value of the Reynolds number.

 

Jacobs figured the effective Reynolds number by multiplying the test Reynolds number by the tunnel's turbulence factor. For the VDT, the turbulence factor was 2.6, the highest of all LMAL tunnels.12 The concept of the effective Reynolds number, though resting on a slender empirical correlation, was soon used by all the NACA wind tunnel sections, in....

 


[
103]

Exterior view of FST, 1930

Loening XSL-1 seaplane in FST, 931.

Fan blades of FST, 1930s.

Above, a view of the Full-Scale Tunnel's huge (434 by 222 feet, and 90 feet high) exterior from the Little Back River, October 1930. Center, one of the first tests in the Full-Scale Tunnel was a performance evaluation of the Loening XSL-1 single engine navy seaplane. October 1931. Below, the FST's enormous twin fan blades.

 

.....particular to show the effects of Reynolds number on maximum lift.13 Some way to compensate for tunnel turbulence was better than no way at all.

The permanent solution Jacobs really wanted to the problem plaguing his work, however, was a new and larger variable-density tunnel with an airstream quality approaching that of the smooth air of free flight. Though.....

 


[
104]

VDT staff, 1929.

The VDT research team, Match 1929. Eastman Jacobs is sitting (far left) at the control panel.

 

....he had wanted it for some time, Jacobs began to campaign in earnest for the new machine after the NACA's introduction in 1934 of its successful 5-digit airfoil series, which had evolved through systematic variation of the nose shape and camber parameters of the better airfoils of the 4-digit family. For the first time since the beginning of the Depression, the Committee was in a relatively good position to secure funds for new construction. In an April 1935 memorandum to the engineer-in-charge reporting the results of a staff conference on ways to increase the speed of airplanes, Jacobs made his idea official. A low-turbulence pressure tunnel, he urged, would greatly enhance the two related lines of research that the VDT team had been long pursuing: development of new airfoils and better understanding of the basic aerodynamic relationship between airstream turbulence, boundary-layer flow, and wing performance. Though asserting that the existing VDT could still provide useful design data and should "probably be maintained for this purpose" and as an air reservoir for the LMAL's 11-inch and 24-inch high speed tunnels, Jacobs quickly emphasized that the "air stream necessary for the continued investigation of the fundamental characteristics of large scale air flows cannot be obtained in the existing tunnel." Turbulence in the old tunnel did not completely invalidate its results for airfoils like those of the 4- and 5-digit classes, but accurate experiments with airfoils and other [105] bodies that might enjoy low-friction laminar flow could not be expected in the existing facility.14

Within two weeks after receiving a copy of Jacobs's proposal for comment, two of Langley's most influential division chiefs sent memos to Henry Reid elaborating their reasons why the NACA should reject Jacobs's idea. Smith J. DeFrance, head of the FST group, questioned whether the knowledge to be gained from the new equipment would warrant the expenditure of money.15 But it was Theodore Theodorsen, head of the small Physical Research Division, who expressed the most vociferous and historically significant (and ultimately incorrect) objections to the facility Jacobs had in mind:

 

I think the variable-density wind tunnels have outlived themselves. I do not think that the variable-density tunnel has led to any fundamental discoveries. They contain a very large amount of turbulence in the airstreams, a condition that cannot be avoided.

 

"What is a new variable-density tunnel to be used for?" Theodorsen asked. "Several years will be required to investigate the tunnel, and then what?" There was "no more need for airfoil testing," the physicist declared, except possibly in connection with some questions about flow conditions in the boundary layer better answered by theoreticians. 16

 

The Jacobs-Theodorsen Rivalry

 

While Eastman Jacobs and his VDT staff had been developing the 4- and 5-digit families using the systematic experimental approach, Theodorsen and his group of more mathematically inclined researchers in the Physical Research Division had been tackling various airfoil problems from the theoretical angle. Though perhaps the greatest contribution of Theodorsen's group during this period was a theory of oscillating airfoils with hinged flaps, related closely to the problem of flutter, the group also provided some very meaningful insight into the relationship between pressure distribution and boundary-layer flow, and hence on wing-section characteristics. In an NACA report published in 1931, Theodorsen had described a "Theory of Wing Sections of Arbitrary Shape," which made it possible, as long as the flow did not separate from the airfoil, to predict the pressure distribution of an airfoil. Starting with an arbitrary airfoil, one changed the closed two-dimensional shape through a conformal transformation almost into a circle; then, by using a rapidly converging series, one transformed the bumpy circle into a true circle about which the flow...

 


[
106]

Eastman N. Jacobs, mid-1930s, and his Sylvanus Albert Reed Award, 1937.

Eastman N. Jacobs, one of Langley's most adventurous, researchers. In 1937 Jacobs received the Sylvanus Albert Reed Award for his contribution to the aerodynamic improvement of airfoils.

 

....was known.17 Though no one at the time thought it reasonable to apply this theory for the purpose of a practical design, the knowledge of the pressure distribution made possible by this clever double transformation later suggested the answer to the riddle of how to shape a laminar-flow airfoil.

The proposed low-turbulence tunnel was not the first issue over which Jacobs, the lab's leading experimentalist, and Theodorsen, the lab's top theoretician, had squared off, and it would not be the last. Beneath the basic difference in their approaches to gaining aeronautical knowledge, there existed a strong personal rivalry and mutual dislike that moved most of their confrontations beyond mere objective disagreement. At Langley both men controlled fiefdoms, and because both men were so valuable to the NACA, George Lewis had permitted the feudal arrangement to flourish. Usually they worked on completely separate activities, but occasionally they had to work together-and then they inevitably clashed.

 


[
107]

Theodore Theodorsen, late 1930s

Theodore Theodorsen, the NACA 's Norwegian import, complained that many LMAL engineers were weak in mathematics.

 

More significant than any hint of personal antagonism in Theodorsen's critique of Jacobs's tunnel proposal was the theoretician's suggestion that Langley's airfoil research had reached an experimental impasse. Though Theodorsen was practical enough to realize that the "imperfect status" of wing theory required designers to make their airfoils "independent of theoretical restrictions," he nonetheless saw the need for the NACA staff to fertilize its experimental routine with a stronger dose of theory. "A large number of investigations are carried out with little regard for the theory," Theodorsen charged, "and much testing of airfoils is done with insufficient knowledge of ultimate possibilities." In his opinion, to discover more advanced airfoils the NACA did not need a new wind tunnel but rather better mathematical and physical understanding of the effects of basic aerodynamic phenomena on wing performance. The implication of his argument was that the experimentalists at Langley had become too interested in and dependent upon equipment for their own good.18

Jacobs disagreed totally with the idea that theoreticians could answer the remaining questions about airfoils better than could experimentalists; he also rejected the argument that it was unnecessary and impossible for the NACA or anyone else to build a pressure tunnel having low airstreams turbulence. He did not disagree, however, with Theodorsen's notion of theory's general role in successful research. An adventurous man with an expansive outlook on what was possible, Jacobs kept up with and understood the most current theory-though he did not devote much of [108] his own time to its study-and valued its role in creating the fundamental but directly useful technological information expected of the NACA. In fact, he had a broader outlook on what was possible than did many of the more theoretical types. During his long career with the NACA (1925-1944), Jacobs explored several revolutionary aeronautical concepts and sometimes grew impatient with co-workers and bureaucrats who saw too many obstacles in the way of their rapid development. In the late 1920s, he tested the potential of thrust augmenters for jet propulsion (see chapter 8). Ten years later, after a newspaper article led him to read the theoretical papers of Hans Bethe, he and fellow NACA researcher Arthur Kantrowitz attempted to initiate the thermonuclear fusion experimentation described in chapter 2. Colleagues remember "Jake" as the type of man who was always looking for the pot of gold at the end of the rainbow. 19

As for the development of airfoils by a combination of theory and experiment, in the 1930s no one working in the United States, perhaps even in the world, surpassed Jacobs's ability. Though it is now difficult to pinpoint just when he first considered controlling the boundary layer through body shape or through control of the usual pressures acting along the body surface, the idea for doing so seems to have been germinating in Jacobs's mind since at least as early as 1930. n a memo on airfoil scale effects in November of that year, Jacobs discussed the importance of the relationship between transition and airfoil drag and mentioned the dependence of the transition point on airfoil shape.20 At the time, he expected that "the possible large drag reductions through prolonging of laminar boundary layers" (that is, through prolonging transition to turbulent flow) would become apparent "as the result of the systematic tests of various airfoil shapes." By 1935, however, he knew this empirical verification would not happen without new turbulence-free testing equipment.21

In May 1935, after considering Jacobs's tunnel proposal together with the comments of DeFrance and Theodorsen, engineer-in-charge Reid determined that the project did "not warrant serious consideration by the Committee at this time." George Lewis, in Washington, concurred.22 The research managers had good reasons to turn down Jacobs's idea for a new VDT. First, other important projects, including the construction of an expensive new tunnel visualized as a super PRT for high-speed propeller research (eventually built at the LMAL as the 19-Foot Pressure Tunnel but which was not really used much for propeller research) were awaiting funding. Second, because the NACA knew that "the desirability of low turbulence in wind tunnels was not widely appreciated," funds for such a facility would be difficult to justify before Congress.23

 

[109] Jacobs Campaigns for Low-Turbulence VDT

 

In late 1935 Jacobs returned to Langley after representing the NACA in Rome at the Fifth Volta Congress on High-Speed Aeronautics. Now more than ever, he was convinced that Langley had to have a low-turbulence pressure tunnel. During his trip he had visited most of the larger aeronautical research laboratories on the Continent, whenever possible examining new experimental facilities and discussing current work. He found the European nations to be in keen scientific and technological competition, spending "large sums of money building up their research establishments." Though concluding that America's "present leading position" in aeronautical research and development was "not seriously menaced at this time," Jacobs warned that "we certainly cannot keep it long if we rest on our laurels" and fail to develop and modernize our test equipment. At the end of his trip report the Langley engineer reverted to the theme of his memorandum of 26 April 1935: "It is again urged that modern variable-density tunnel equipment be built in this country capable of testing at full dynamic scale for modern aircraft."24

Jacobs also brought back some new insight into the nature of the boundary layer. While in England he had spent a weekend at the home of Sir Geoffrey I. Taylor, professor of physics at Cambridge University, who had presented a paper on high-speed flow at the Volta Congress. In long private conversations, Taylor described for Jacobs the substance of his recent work in the statistical theory of turbulence. This theory seemed to indicate that "the transition from laminar to turbulent flow was due to local separation caused by the pressure field."25 By implication, this result said that transition could possibly be delayed or perhaps avoided by preventing laminar separation-i.e. by using a falling pressure gradient. This would be the mechanism used eventually by Jacobs in his design of laminar-flow airfoils.

Jacobs also had the chance at Cambridge to talk at length with B. Melville Jones, professor of aeronautical engineering, who, like Jacobs, epitomized the researcher who combined theory and experiment for practical purposes. (Jones's classic 1929 paper, "The Streamline Airplane," had provided designers with an idealized goal that served to indicate how much power was being wasted to overcome drag.) Jones reported to his American counterpart that recent British flight work showed considerable laminar flow over the forward regions of very smooth wings where much of the flow was in the falling-pressure region. This encouraged Jacobs greatly; It pointed to the possibility that drag levels achieved by well-designed [110] advanced aircraft could be down to the value of skin friction. Thus, the only remaining opportunity for reducing drag would lie in encouraging laminar flow-something that is still true. 26

Armed with this new information, Jacobs returned to the LMAL ready to press harder for the construction of his new variable-density tunnel. During his presentation to the NACA's annual manufacturers' conference in May 1936, he advertised his belief that further reduction in drag would have to take place as a result of somehow delaying transition to turbulent flow in an airfoil's boundary layer.27 At a laboratory conference on boundary-layer control in July, Jacobs argued that "direct control through shape should be placed first on our program" and again urged his colleagues to support his idea for the construction of suitable turbulence-free testing equipment.28 In the fall, he wrote a paper on "Laminar and Turbulent Boundary Layer as Affecting Practical Aerodynamics," which, in essence, was a plea for the new tunnel.29

On 28 May 1937, Jacobs's 13-month campaign for a low-turbulence VDT finally achieved its goal, if in a roundabout way: NACA headquarters authorized the construction of an "icing tunnel." The name was a necessary political subterfuge. George Lewis felt that the NACA could not at the time justify the expense of a new wind tunnel at Langley solely for development of low turbulence. Congressmen simply would not understand the urgency. But the Committee could sell it, he believed, on the basis of icing experiments. Many aircraft crashes traced to icing problems were attracting considerable public attention in 1937; the commercial airline operators were clamoring for useful information on the subject.30 Here was a way for the NACA to kill two birds with one stone.

In 1937 a team of Langley researchers headed by Lewis A. Rodert was in fact in the midst of conducting icing research in free flight. The idea was to pipe hot engine exhaust gases through interior passages in model wings mounted firmly on struts a foot or two above the wing of a test airplane-at the critical altitude where air temperature could cause ice to form. When the plane reached that height, water was sprayed on the leading edge of the model. As the edge quickly coated with ice, heat was piped into the model's interior passages, and a timed camera recorded how long a given amount of heat took to free the surface of its ice coating. This technique worked-that is, it worked in these flight tests in small models. But it raised serious problems of adaptation for full-size flying machines. In particular, since the heat-conducting pipes in an actual airplane had to pass through critical elements of the wing structure (e.g., spars and ribs), the technique threatened to weaken that structure seriously, mainly by adding too much weight. Thus tests using models in a small ice tunnel could not aid the flight [111] program significantly. (In 1938 Rodert's team reduced the weight of the NACA thermal ice-prevention system to a minimum. The army provided funds for Langley to remodel a Lockheed 12 with a wing-and-tail heating system and to send the aircraft up into the clouds seeking ice.)31

So Langley never really intended to conduct many icing experiments in the icing tunnel of 1937. LMAL technicians had insulated the walls on the outside of the tunnel with only a wrapping of crude insulation (kapok removed from surplus navy life preservers by members of the Hampton High School football team) and added only the simplest refrigerating equipment (an open tank of ethylene glycol cooled by blocks of dry ice). On one hot summer day in 1938, when everything was ready, enough of the cold mixture was pumped into the tunnel test section to cause some ice to form on the leading edge of an airfoil. Then a method of using an engine's exhaust heat to prevent the ice formation was tested. A perfunctory series of experiments fulfilled the announced purpose of the ice tunnel, and Langley immediately converted it into a low-turbulence tunnel for low-drag airfoil studies. This facility served as a prototype for the pressurized Two-Dimensional Low-Turbulence Tunnel constructed at the LMAL between 1939 and 1941.

 

In Search of the Laminar-Flow Airfoil

 

After returning from the Volta Congress in late 1935, Jacobs discussed with the men of his VDT section what he had discussed in England with Taylor and Jones: the idea that continuously decreasing pressure along an airfoil would tend to delay transition from laminar to turbulent flow in the boundary layer. But exactly how and when the implications of this concept were first spelled out for the rest of the lab is not entirely clear. Jacobs now recalls that sometime in 1937 he "rediscovered" the idea, noting that the effect of the pressure gradient on laminar separation had been established previously.32 In fact, the idea dated back to a paper by Prandtl published before World War I, and had been restated by Theodorsen in his 1931 paper, "Theory of Airfoils of Arbitrary Shape." Whatever its origins, this concept was the underlying criterion for the Jacobs group's imminent preliminary design of laminar-flow airfoils.

Jacobs and his fellow researchers knew that laminar-flow airfoils would have to satisfy several conflicting requirements. They were confident that an application of existing airfoil theory could start the design process by producing shapes with prescribed pressure distributions. (Langley's airfoil experts already had some valuable experience in designing airfoils for a prescribed pressure distribution. In the mid-1930s they had designed the 16-series to have a specific distribution in order to achieve the highest.....

 


[
112]

In late September 1987 Langley performed stalling and icing studies with a DC-3 Mainliner passenger transport

.

In order to warn the pilot of an approaching stall, the NACA engineers installed sharp leading edges on the section of the wing between the engine and fuselage (left)

.

In order to simulate the effects of ice formation on the DC-3's performance, the engineers cemented pieces of sponge rubber to the forward part of the wings (right)

In late September 1987 Langley performed stalling and icing studies with a DC-3 Mainliner passenger transport (top) belonging to United Airlines. In order to warn the pilot of an approaching stall, the NACA engineers installed sharp leading edges on the section of the wing between the engine and fuselage (left). These sharp edges disturbed the airflow enough to cause a tail buffeting, which could be felt by the pilot in his control column. When the pilot felt this buffeting, he knew that his airplane was approaching a stall and needed pilot correction. In order to simulate the effects of ice formation on the DC-3's performance, the engineers cemented pieces of sponge rubber to the forward part of the wings (right), where ice was thought to form most often, and then measured the resulting changes in the plane's climb, cruise, and stalling speeds.

 

[113] .....possible critical Mach number.)33 But they were also quite sure that such a theoretical application would only be the first step. Theory alone could not answer the key design questions, such as: What distribution was needed for laminar flow? What compromises with other kinds of design requirements would have to be made for construction of an effective and practical wing? Answers to these questions would have to be found, the researchers believed, in a comprehensive program of experiments in the low-turbulence tunnel.

The Jacobs group could visualize a virtually infinite number of airfoils with falling pressure distributions-with varying pressure gradients, camber, thickness, and positions of peak suction pressure. A large number of related experimental airfoils would now have to be designed to incorporate the falling pressure feature, together with systematic variations in the other parameters.34 Obviously this would be a far more difficult task than had been the design of the previous NACA airfoil families, which involved mostly simple, arbitrary, geometrical relationships.

With his commitment to the design of laminar-flow airfoils now overshadowing all of his other work, Jacobs disappeared from Langley Field for a few days to unravel the mysteries of Theodorsen's 1931 airfoil theory and to explore possible ways of reversing its procedure, which had been designed to predict the pressure distribution from a given shape. First, he called over to his house a close friend, Robert T. Jones, a highly intuitive NACA researcher who had taken a few classes at Catholic University taught by Max Munk.35 Together, Jacobs and Jones decided that Theodorsen's method could not be used in the way desired without adding to the theory. Jones proposed an extension of the theory derived from Munk's thin-airfoil work that seemed to be a way of calculating a shape that would give a desired sequence of pressures, but this also proved too inaccurate.36

When Jacobs returned to the laboratory from his short working vacation, he challenged his staff to apply Theodorsen's theory in design. H. Julian "Harvey" Allen, one of the brightest members of the VDT staff, came up with one means of inverting the theory based on a linearization that started from a thin Joukowski airfoil. Applicable only to thin sections, Allen's way proved too inaccurate near the leading edge for prediction of local pressure gradients.37

No one in the VDT section had any special training in advanced mathematics of the sort required, which prompted a few of the men to approach Theodorsen's Physical Research Division for assistance. According to Ira Abbott, another key member of Jacobs's staff: "We were told that even the statement of the problem was mathematical nonsense with the implication that it was only our ignorance that encourages us."38 Theodorsen himself went to the trouble of showing that the shapes likely to result from an [114] inversion of his theoretical method would be "unreal," things that looked like figure eights and surfaces that crossed over one another.39 Encouraged now by hearing this negative pee response, Jacobs stubbornly persisted in directing an all-out effort to devise a satisfactory inversion of the Theodorsen method. (In fairness to Theodosen, it must be noted that he eventually contributed to solving the problem)40

The breakthrough in this effort came in the spring of 1938, during the construction of the icing tunnel The inversion, which Jacobs now says he modeled after Isaac Newton's clever method of approximating a square root, consisted essentially of charging a function in small increments in the conformal transformation of Theodorsen's theory. By taking an ordinary wing section, like the N.A.C.A. 12, and "running it backwards," that is, designing its nose features according to the shape principles of the tail and its tail features according to the nose, Jacobs's team was able to arrive at an approximate shape that had falling pressures over most of the surface.41 It is impossible to document whether this single spectacular approximation ever took place; the inversion procedure may in fact have been a gradual refinement. Jacobs's role is not its dispute, however; he was the inspiration and driving force behind the entire laminar-flow program.

After verifying its pressure distribution theoretically, Jacobs rushed the manufacture of a wind tunnel model through the LMAL shop. As soon as the new low-turbulence testing equipment was ready for operation, he supervised a test of the new model in comparison with a conventional airfoil. To his delight, "the new airfoil showed a drag on the order of one-half that of the conventional airfoil." 42 The result pleased him for two reasons especially: it provided empirical verification that inversion of the Theodorsen theory worked-something that his rival Theodorsen had called impossible-and it further justified the construction of the controversial new VDT, which he had personally championed through strong opposition.

Inspired by this success, Jacobs and colleagues explored further into the range of shapes theoretically enjoying laminar flows. By combining experimental knowledge with better ways of approximating solutions, they delineated a family of airfoils designated the 2-, 3-, 4-, and 5-series airfoils. Wind tunnel and free-flight work on some of these sections provided good qualitative information about the characteristics to be desired, but, because the mathematics was simply too approximate to show correctly the effects of changing such key parameters the profile of the section near the leading edge, the work produced no practical airfoils. Much was learned, however,....

 


[
115]

Historical evolution of airfoil sections, 1908-1944.

The historical evolution of airfoil sections, 1908 1944. The last two shapes (N.A.C.A. 661 -212 and N.A.C.A. 74 7A315) are low-drag sections designed to have laminar flow over 60 to 70 percent of chord on both the upper and the lower surface. Note that the laminar flow sections are thickest near the center of their chords.

 

 

[116] .....and modified criteria evolved or development of the new 6 series.* In comparison with conventional sections, airfoils from this new series had the point of maximum thickness further aft along the chord. The point was prescribed in order to achieve the type of pressure distribution on the airfoil surface thought necessary for laminar flow.

In June 1939, the NACA distributed an advance confidential report prepared by Jacobs covering his new laminar-flow airfoils and explaining the methods his wind tunnel team had adopted for airfoil and boundary layer investigations. 43 Though he Committee did not circulate the exact results of this research publicly until after World War II, a copy of the confidential report was sent to the Paris office of the NACA; John Ide burned it along with all of his other files before the Germans overran the city in 1940. German aeronautical engineers had reason to guess at the nature of the development anyway. On the first page of its Annual Report for 1939, published in 1940, the NACA hinted:

 

Discovery during the past year o a new principle in airplane-wing design may prove of great importance. The transition from laminar to turbulent flow over a wing was so delayed as to reduce the profile drag, or basic air resistance, by approximately two-thirds.

 

Though admitting that it was still too early to appraise adequately the significance of this achievement the NACA nonetheless suggested that its continued wing research should in the near future "increase the range and greatly improve the economy" o both military and commercial aircraft.

Beginning in 1940, Langley y helped North American Aviation test fly its prototype of the P-51 mustang, the first aircraft to employ the NACA laminar-flow airfoil. ** Though the Mustang's war record confirmed....

 


[
117]

In the spring of 1941 Langley installed an experimental low-drag test panel on the wing of a Douglas B-18 airplane.

The panel, seen close up at right, was fitted with suction slots and pressure tubes for a free-flight investigation of the transition from laminar to turbulent flow in the boundary layer

In the spring of 1941 Langley installed an experimental low-drag test panel on the wing of a Douglas B-18 airplane. The panel, seen close up at right, was fitted with suction slots and pressure tubes for a free-flight investigation of the transition from laminar to turbulent flow in the boundary layer. The pressure at each tube was measured by liquid manometers installed in the fuselage.

 

.....expectations of appreciable improvements in speed and range as a result of the low-drag design, practical experience with this and other aircraft using advanced NACA sections in the 1940s also showed that the airfoil did not perform quite as spectacularly in flight as in the laboratory. Manufacturing tolerances were off far enough, and maintenance of wing surfaces in the field careless enough, that some significant points of aerodynamic similarity between the operational airfoil and the accurate, highly polished, and smooth model that had been tested in the controlled environment of the wind tunnel were lost.44 Still, despite manufacturing irregularities and the detrimental effects of actual use, the Mustang's modified 4-series section, with its pressure distributions and other features, proved an excellent high speed airfoil. The delineation of it and other laminar-flow airfoils was thus a great contribution by Langley, even if not exactly to the degree advertised by NACA publicists like George Gray, who claimed in Frontiers of Flight that "the shape of this new wing permitted the flow to remain laminar until the air had traveled about half way along the chord." According to Langley engineers who knew what it took in practice to achieve success, Gray's claim was an exaggeration. Because the percentage drag effect of even minor wing surface roughness or dirt increased as airfoils became more efficient, laminar flow could be maintained in actual flight operation only in a very small region near the leading edge of the wing. 45

Though the NACA's high-speed airfoil work continued to be impressive in the late 1940s and early 1950s, this chapter's examination of its role in...

 


[
118]

The North American XP-51 Mustang (shown above, in 1948)

.

In 1946, Langley equipped a P-51B (left) with wing gloves for an investigation of low-drag performance in flight.

.

The North American XP-51 Mustang (shown above, in 1948) was the first aircraft to incorporate an NACA laminar flow airfoil. In 1946, Langley equipped a P-51B (left) with wing gloves for an investigation of low-drag performance in flight.


 

....the history of the VDT actually ends in 1939 with the cautious public announcement of the laminar-flow airfoil, a dramatic research success. By that time, the Committee's airfoil research had moved a full 180 degrees away from its unsatisfactory course of 20 years earlier, when a very small research staff with very limited technical capability and no operational test facilities of its own had mainly occupied itself accumulating, analyzing, and disseminating European data. Thanks in large part to the VDT and to its enhanced successor, the Two-Dimensional Low-Turbulence Tunnel, a much larger research staff worked at the cutting edge of modern experimental technology by the time of American entry into World War II. One result: wing sections developed by the NACA at Langley became by far the most widely used sections worldwide.

 

[119] Postscript to VDT History

 

The history of Langley's variable-density wind tunnel would not be complete without some reference to the ultimate fate of the actual equipment. After replacing it with the two -Dimensional Low-Turbulence Tunnel in the early 1940s, Langley continued to use the VDT, minus internal test instruments and mechanisms, as a high-pressure air storage tank for small high-speed induction tunnels. In 1981 the Pressure Systems Committee of NASA Langley Research Center closed the VDT pending its inspection and recertification for use as a tank, but Langley lacked the $30,000 needed to ready the old, riveted structure for inspection. In 1984 the National Park Service recommended that step be taken to safeguard the VDT as a national historical landmark.

The fate of Max Munk, the father of the Langley VDT, should also be noted briefly. After leaving the lab n 1927, Munk seems to have "failed to repeat the brilliant record, 46 he achieved when the VDT and other NACA resources were available to him. He took a job with Westinghouse in Pittsburgh, where he tried to solve a cooling problem in electric motors. Then he worked a year for the American Brown Boveri Electric Corporation of Camden, New Jersey, and another year for the small Alexander Airplane Company in Colorado. In the late 1920s Munk asked the NACA to publish one of his articles, but the Committee rejected it for lacking clarity and rigor. By 1930 hard times had reduced him to writing "pathetic letters"47 in which he styled himself "the foremost aerodynamic expert of the world" and declared that "all special scientific methods by which aircraft is [sic] computed nowadays, most experimental methods, and types of equipment have been originated by me."48 During the Depression he became a consulting editor for the journal Aero Digest, and, in the opinions of George Lewis and others at NACA headquarters, contributed anonymously to its editorial campaign against the Committee.49 Munk also taught mechanical engineering a Catholic University part time and educated himself in patent law.

It is not widely known that Munk proposed to design another new wind tunnel for the NACA. In July 1939, as Nazi Germany prepared to invade Poland, Munk wrote a letter to NACA chairman Joseph Ames, the man who had arranged for Munk to come to America in 1920 and work for the NACA, saying that he knew how to design "the ideal, the most efficient, most practical, most useful and most impressive piece of equipment" for the study of high-speed airplane problems He suggested that the NACA might make use of his knowledge by rehiring him as an employee or by special contract. 50

[120] NACA member Jerome C Hunsaker, the other principal actor in Munk's immigration to America, answered Munk for Ames, who was very sick, with a one-sentence letter: "I have read your letter to Dr. Ames about a proposed wind tunnel, but unless you can disclose something of the ideas behind it, I don't see how anything can be done about it." Two weeks later George Lewis wrote Munk, suggesting that he "submit for the Committee's consideration general or detailed plans" of the proposed device.51

Munk swallowed hard, for he had had a very serious falling-out with Lewis in 1927, and sent the director of research a contrite letter in which he proposed in vague terms the design of a "new type of wind tunnel," the same phrase Munk had used in 1921 to describe the VDT. This new tunnel was to be at least 32 feet in diameter t the throat and have a 20,000-horsepower motor capable of providing 400-mile-per-hour low-turbulence airflow. Munk estimated its cost at $1.5 million.52

Lewis, pressed by the heavy schedule of preparing the expanding NACA organization for wartime research and development activities, was slow to act on Munk's proposal.53 Eventually he did ask Langley's foremost designer of large atmospheric wind tunnels, Smith DeFrance, for a quick appraisal of Munk's idea, to be based on the correspondence. DeFrance reported back that a device of the size and speed suggested could not attain 400 miles per hour with only 20,000 horsepower, and probably would not be of much value to the Committee even if it could. "It is apparent that Dr. Munk has in mind testing single-engine pursuit ships at full-scale and at what he may consider to be full speed," be France asserted. "However," he went on, "from experience at Langley Field, it is safe to say that results obtained at 250 MPH can be extrapolated to 400 MPH, provided compressibility effects are disregarded." As for determining compressibility effects, DeFrance argued that either of the two new 16-foot wind tunnels authorized for construction by the NACA in 139 (one at Langley and the other at the new Ames laboratory in California) would supply the necessary information.54

Munk later submitted a more formal and specific contract proposal for the design of his new tunnel, out in May 1940 the NACA rejected it.55 Munk then asked Vannevar Bush, who had replaced the ailing Ames as NACA chairman in October 1939, to reestablish his old technical assistant position at NACA headquarters and to appoint him to it. Bush, after looking into the NACA's past problems with the imported aerodynamicist, turned down that idea as well.56

Munk had to remain content writing articles for Aero Digest and teaching part time at Catholic University. Beginning in 1945, he went to work as a research physicist at the Naval Ordnance Laboratory, contributing reports on the mechanism of turbulent fluid motion. He returned to Catholic....

 


[
121]

Max Munk at his home in Rehoboth Beach, Delaware, 1981.

Max Munk at his home in Rehoboth Beach, Delaware, 1981.

 

.....full time in 1958, retiring in 1961. In the mid-1970s, the American Institute of Aeronautics and Astronautics (AIAA) honored Munk with one of its awards. In 1977, Munk published a small book at his own expense in which he claims to have provided the proof of Pierre de Fermat's "Last Theorem," which has baffled the mathematical profession for over 300 years.57 In 1985, 95-year-old Max Munk was still living and still in good health, if with failing eyesight, in Rehoboth Beach, Delaware. He enjoyed discussing with visitors the etymology of words, especially Greek derivatives, and quoted at length from the works of Arthur Schopenhauer.

The consequences of the NACA's first and later rejection of Munk on the quality and vision of subsequent research at Langley laboratory are still a matter of debate. Basic work in theory seems to have declined for a while at Langley following Munk's departure; over the years there would be few researchers at the NACA who spoke the language of higher mathematics. On the other hand, the overall quality of NACA research in the 1930s seems not to have declined but, in fact, to have risen.

There is far less contention about the consequences for Munk himself: the impact of the revolt on Munk's life and career after 1927 was tragic. His notorious departure from the NACA surely slowed the advance of his ideas within the American community of aeronautical engineers. The NACA, which so often in the early 1920s had touted Munk, as it's most brilliant and productive staff member, now treated him virtually as a no person. Many technical reports published by the NACA that by rights should have [122] credited Munk's earlier papers cud not reference them at all. As a result, many members of the American aeronautics community supposed that Munk's work was irrelevant or out-of-date. They began to assume (perhaps correctly) that Munk had pretty well exhausted his supply of genius and vision by the time he left Langley. Moreover, they suspected from what they did hear from and about him that he was devoting far too much of his subsequent time and energies to flirting with exotic research topics (such as the Flettner rotor, a strange sailboat-like craft moved by vertical rotating sheet-metal cylinders-a concept which also interested Albert Einstein and Jacques Cousteau) and criticizing the research establishment for not investigating their potential benefits.

Perhaps the bitterness with which Munk remembered the revolt against him at Langley made him think he had something new to prove. Perhaps the hurt and anger did affect Munks work adversely, if by adversely one means that by choosing to explore research problems offering largely imaginary benefits instead of those having the most urgent technological relevance, Munk failed to match the practical brilliance of his NACA contributions. In spite of all the reasons he might have had for holding a grudge against the NACA, Munk might have risen above them further than he did. As Samuel Johnson once said, "A man of genius is seldom ruined but by himself." On the other hand, America with its egalitarian society-with its egalitarian engineering society-is not an easy place in which to be a genius.

 


* Airfoils belonging to the 6-series were designated by a six-digit code together with a numerical expression of the type o mean line used. For example, in the designation "N.A.C.A. 65,3-218, a = 0.5." 6 was the series designation; 5 denoted the chord wise position of minimum pressure in tenths of the chord behind the leading edge for the basic symmetrical section at zero lift; 3 was the range of lift, coefficient in tenths above and below the design lift coefficient for which favorable pressure gradients existed on both surfaces; 2 was the design lift coefficient in tenths; 18 indicated the airfoil thickness in percent of the chord; and a = 0.5 showed the type of mean line used. When the mean-line designation was not given after the sixth digit, a uniform-load mean line ( = 1.0) had been used. Ira H. Abbott, Albert E. von Doenhoff, and Louis B. Stivers, Jr., "Summary of Airfoil Data," TR 824, 1945.

**The P-51 had another interesting distinction: it was the first case of an aircraft's actual construction matching its aerodynamic design specifications without adding thickness in building the metal skin. The idea of the Mustang designer, a German perfectionist named Edgar Schmued, was to produce an airplane whose aerodynamic shape was the same as that decided upon by the aerodynamicist not that shape plus an overcoat of lapped aluminum alloy that in places might add up to four sheets of thickness. The Mustang's faithfulness to profile was later exceeded by refined thicker-skinned aircraft like the Lockheed P-80 and F-104. See Richard P. Hallion, Designers and Test Pilots (Time-Life Books, 1983), pp. 78-79, 148-151, and Richard Sanders Allen, Revolution in the Sky: Those Fabulous Lockheeds (The Stephen Greene Press, 1967).


previous pageindexnext page