Experimental Aircraft

Walter J. Boyne

OVER the years, the style and character of experimental aircraft, and particularly their use, have changed fundamentally, to the great advantage of science and to the loss of romance. To understand the status of experimental aircraft today and anticipate possible outcomes in future efforts, it might be worthwhile to retrace the development of a few experimental aircraft of the past.

In the beginning, all aircraft were experimental in one sense or the other; however, of the beginners, only the Wright brothers’ aircraft were experimental in a scientific sense. In terms of aeronautics, the Wrights were in advance of all others by a minimum of six years; in procedural experimental terms, they were in advance of all others by a decade.

The Wrights, without a formal education, but self-educated with wonderful discrimination, plunged into an arduous, often disheartening three years of planning and testing that led from their experiments with kites to the flawless execution of four flights on 17 December 1903. Not surprisingly, these successful first flights were not widely known and, where known, were often discounted. The Wrights continued with their experimentation for the next two years, “perfecting” their design by 1905 and then retiring from flying for two years to sell their invention while protecting their patents.

In Europe, there was nothing to correspond to the Wrights’ insightful and rapid development program. The news that man had flown was not believed, and the pioneers–from Santos Dumont through Voisin and Ferber–sought flight through intuition rather than many would not agree with it today. But the facts are clear: no systematic, step-by-step approach, allied to a fateful insight, was achieved on the continent. Not until the demonstrations at Reims and elsewhere in Europe established the fundamental Wright baseline did the method of development by intuition achieve success. The European effort, spurred by military expenditures, soon eclipsed all American efforts, but it did so on the foundation of the original Wright experimentation and success.

Since the Wrights, more than a million aircraft have been produced in countries all over the world. Thousands of individual types have come and gone, many not remembered or even recorded by drawing or photograph. In this process, the U.S. Air Force and its predecessor organizations have contributed a host of remarkable experimental aircraft, which reflect not only the technology of the times but the spirit and rigor with which experiments were conducted.

The United States, after having invented the airplane, promptly forgot about it, although the U.S. Armed Forces had observers at the front long before America’s involvement in World War I, and the airplane had become headline news in the newspapers of the world. The warring nations had initiated conflict with a few aircraft relegated to ancillary duties, all of which were basically derivatives of the ad hoc intuitive development of aircraft for prewar sportsman pilots. The tempering experience of war created an enormous industry (England produced more than 55,000 airplanes in the First World War; Germany, more than 40,000) together with a series of disciplines that remain with us to this day. It is not generally recognized, but within the first nine months of combat in World War I, almost every aspect of modern aerial warfare had been demonstrated, including strategic bombardment (the Avro 504 raids on the Zeppelin sheds), psychological warfare (the Taube’s bombardment of Paris), strategic reconnaissance (the monitoring of General Alexander von Kluck’s curving arc above Paris), ground attack, aerial photography, and even air-to-air combat. By March 1915, things had progressed to the state that an entire battle, Neuve-Chapelle, had been fought on the basis of maps prepared from aerial photography and in conjunction with raids to interdict rail lines.

Every air service established an experimental station: the British at Farnborough, the French at Meudon, the Germans at Johannisthal, and the United States at McCook Field (Dayton, Ohio).

The Americans were at an initial disadvantage, beginning the war with some fifty-five obsolete training planes and making the logical but costly decision to commence production of established Allied types, including the English de Havilland DH-4, the Handley Page 0/400 bomber, and the Italian Caproni.

In a manner that became characteristic of U.S. air endeavors, McCook Field became a focal point, a collecting agency, for some of the brightest young flyers in the business, as well as the most talented engineers. They took from their foreign colleagues and applied to it a work discipline that resulted in the creation of engineering logistic and test entities which led directly to today’s Air Force Systems Command and Air Force Logistics Command.

The experimental process took time to mature, however, and much was vested in the pilot’s almost intuitive analysis. It is interesting to read McCook Field test reports today: some are as much as eight pages long, filled with succinct comments such as “good ship,” “controls need work,” or “please condemn.” There were extensive tests underlying the pilot’s analysis, many of which would be familiar today, but in the main, a pilot could make or break the development of an aircraft with his comments.

Among the literally thousands of aircraft that have followed the experimental path, I shall discuss a number of those that might not be the most famous of their kind but which illustrate aspects of the experimental process that might not otherwise be considered.

the Verville Sperry R-3

The Verville Sperry R-3 is an almost perfect example of the opportunity cost of an inadequate development program. Designed by Alfred Verville, a kindly genius who had a penchant for just missing the brass ring of commercial success, the R-3 was years ahead of its time when it first appeared in 1922 as a certain winner for the Pulitzer Trophy Race.

Here was a racer, contemporary with the Thomas-Morse biplane pursuit, which featured a cantilever wing, streamlined fuselage, and fully retractable landing gear, clearly presaging the mid-1930s formula of the Messerschmitt, Hurricane, and Spitfire. However, it also evoked some political problems that might be analogous to the current F-16/F-20 controversy. The Verville was developed by the McCook Field Engineering Division and manufactured by the Lawrence Sperry Aircraft Company of Farmingdale, New York. Three aircraft were purchased, and on them, the aircraft builders intended to use the silky-smooth 450-horsepower Curtiss D-12 engine and the all-metal Curtiss Reed propeller. Fundamental to the design was the use of the patented Curtiss wing radiators, thin brass sheets that conformed to the airfoil.

It happened that the foremost aircraft manufacturer of the time was the Curtiss Aeroplane and Motor Company, which was also building the sleek series of racing biplanes. As a political result, the R-3 was fitted with the 300-horsepower Wright H-3 engine, notorious for its vibration. A stock wooden propeller and “lobster pot” Lamblin radiators were installed. With these totally undesirable modifications, the airplanes were no longer competitive, and first and second pieces were won by the sleek Curtiss biplanes using the preferred engine/propeller/radiator combination.

All three R-3s started the race, but only two finished. Lieutenant Eugene Barksdale finished fifth at a little better than 181 mph. Lieutenant Fonda B. Johnson finished seventh, his engine freezing solid immediately after landing. The legendary Lieutenant Saint Clair Street broke an oil line and had a forced landing, damaging the airplane.

Development of the aircraft ceased for all practical purposes, despite the large investment. There were several problems with it–incipient flutter, the drag induced by the open wells of the retracted wheels, a general lack of harmony in the controls–that would have been eliminated by a series of tweaking test flights or in the wind tunnel. For political and economic reasons, these remedial procedures were denied.

A Curtiss D-12 engine was installed in the plane or the 1923 Pulitzer, and while vibration was no longer a problem, there were still handling difficulties, especially at top speed, now reaching 233 mph, The airplane had to withdraw from the race. Once again, a Curtiss biplane was the winner.

Again, no substantial development work was invested in the design, and it was with some misgivings resurrected for the 1924 Pulitzer, when the preferred entry–a Curtiss biplane–crashed. Ironically, the R-3, piloted by Lieutenant Harry H. Mills, won the race at a slow speed of 215 mph. The racer was almost immediately relegated to the McCook Field Museum, where it was ultimately burned. The R-3 remained merely another exciting, unfulfilled concept.

Huff Daland LB-1

Sometimes the experimenters almost got things right, only to be frustrated by an outside event. In the case of the Huff Daland LB-1, the intended replacement for the series of Martin Bomber-inspired MB-2, which formed the bulk of the bomber fleet, engine reliability was of such a low order in the early 1920s–and for a considerable period there-after–that a twin-engine aircraft was far more susceptible to crashing from engine failure than a single-engine type. The reason, of course, was that no twin-engine aircraft of the period could maintain flight on a single engine, so by running two engines you doubled the probability of an in-flight emergency.

Huff Daland built a giant–well, sixty-six foot wingspan–single-engine bomber to avoid the redundant emergency problem, using the Packard 2A-2540 engine of 750 horsepower. The Packard, like all of the Packard aviation engines of the time, was a monument to unreliability; and it did not take long to determine that one Packard would fail more often than two Liberties. In this instance, however, Huff Daland retrieved the situation by falling back on the twin Liberty engine formula for the LB-5, which led, in turn, to the whole series of Keystone bombers that served as the background of the fleet until the early 1930s. Here, failure led to success on an unpremeditated scale.

Boeing XP-9

The Air Corps was increasingly interested in monoplanes in the late 1920s, and the first Boeing effort in this regard was their Model 96, the XP-9. This aircraft actually proved to be more important for structural than configuration reasons, for it was also the first metal semi-monocoque fuselage by Boeing, and it was to have great influence on a number of later designs.

The XP-9 was powered by the standard liquid-cooled Curtiss SV 1570 engine of 600 horsepower, and some sources indicate that it had a top speed of 213 mph. Its biplane counterparts, the Boeing P12D and Curtiss P-6E, had top speeds of 188 and 193 mph, respectively.

The XP-9’s handling characteristics and landing speeds left much to be desired, however, and no production order ensued. Yet the airplane’s influence was far greater than commonly realized, for it inspired the Boeing Monomail, a single-engine, all-metal, retractable-gear mail plane, and the YB-9 Death Angel. The Death Angel, in turn, pioneered the construction that resulted in the precedent-shattering Boeing 247D transport and led directly to the Model 299 Flying Fortress. Despite the test pilot’s report, which called the aircraft “a menace” because of its poor visibility and bad flying qualities, the XP-9 had an influence well beyond the manufacturer’s expectations.

Fokker YO-27/XB-8

Often manufacturers have experimented in a most economical way, stretching existing technology to cover new configurations. The result has rarely been satisfactory. The American Fokker company had inherited the design philosophy and manufacturing techniques of the parent Dutch Fokker company; and these in turn, extended back to the wartime work of A. H. G. Fokker and Reinhold Platz. Fokkers were built with steel-tube fuselages and wooden wings, and they would be so until after Fokker’s death at the age of forty-nine in 1939. The Air Corps issued a call for a monoplane light bomber and/or observation plane, and the Fokker firm responded to the new configuration with their familiar construction techniques, adding only a retractable landing gear as a token to modernization. The XB-8 was, in fact, the first retractable-gear bomber to reach Wright Field.

The airplane was in direct competition with two versions of a basic Douglas design, the XO-35 and the XB-7. On balance, the Fokker was a cleaner aircraft but otherwise very similar in terms of weight, wing area, and engines. The Douglas airplanes, however, were all metal, had a more modern airfoil, and generally performed much better than their Fokker competitor. At least ten miles per hour faster, the Douglas airplanes were also much more pleasant to fly.

Fokker had erred in pursuing the same formula: the wooden wing, with its standard Fokker airfoil and layout, simply was not suitable any longer. The traditional thick section created too much drag, and the Fokkers were as much as twenty miles per hour slower than the Douglas YB-7, which topped out at 182 mph. Both companies received token orders, but the stage had been set for the first truly modern bomber, the descendant of the XP-9, the Boeing YB-9.

the Boeing YB-9

The YB-9 bomber was in the direct development line that led to the B-17, and it was revolutionary. It combined a cantilever wing and retractable landing gear with a blistering 173-mph top speed, more than 50 percent faster than the Keystones in fleet service. It was a private venture by Boeing and looked like a world beater. Unfortunately, Wright Field had been working for an extended period with Martin and developed in concert the XB-907, which gained the only production orders. Only seven models of the YB-9 type were built, but they led directly to the world-beating 247D transport, which, in turn, laid the foundation for the B-17 and Boeing’s forty-year dominance of the bomber industry.

Martin XB-907 (B-10/B-12)

In these days when castigating the Department of Defense is so popular, one can look back to the 1930s with some nostalgia. Then the Army and Navy were revered institutions, but they didn’t always get their proper share of credit. The McArthur-Pratt agreement was formalized on 9 January 1931, assigning to the Air Corps the air defense of coastal regions. The Air Corps put out a request for proposal on an aircraft that would replace the traditional coastal defense weapons, asking for an advanced all-metal monoplane. The Martin Company responded with proposals for a biplane and a monoplane with fixed gear, traditional biplanes. There ensued an agonizing two-year process in which Wright Field asked for an all-metal monoplane, retractable landing gear, enclosed cockpits, and cantilever wing; Martin resisted each of these efforts but finally succumbed to pressure to create the XB-907, which incorporated the retractable gear and wing structure developed at Wright Field. The prototype had some problems but, with further help from Wright Field, developed into the XB-907 (later the XB-l0), with a turret and a 207-mph top speed, taster than any service fighter.

The B-10/B-12 series taught the pilots, bombardiers, and mechanics both what a modern airplane could do and what it required in terms of maintenance and support, laying the foundation for the great fleets of B-17s that would follow.

Curtiss A-14

Sometimes nothing happens even when everything goes well. Curtiss was notorious for stretching designs long past the point of no return–the basic PW-8 design had appeared in 1924 and been tweaked for the next ten years before ending life as the export Hawk III and IV. Similar life extension had been provided the Falcon series of observation planes.

With 1934’s XA-14, however, Curtiss broke entirely new ground. The XA-14 was a strikingly handsome all-metal, twin-engine attack plane, fitted at one time with a 37-mm cannon. Top speed was a sizzling 254 mph, and it was reportedly delightful to fly. However, it was three times as expensive as the Northrop A-17A just coming into service; and Congress, as Hitler was reputed to be, was more concerned about numbers than performance. The XA-14 was developed into the attractive A-18, but only thirteen of these were ordered.

In this instance, the experimental aircraft did not come into service, nor did it have much effect on the Curtiss firm’s thinking. However, it did spur the Air Corps planners to raise the requirements for attack planes to a European level and to promote a competition that resulted in the Douglas DB-7–progenitor of the A-20 series.

Stearman XA-21

The prospect of war inflamed the imagination of designers, and companies recklessly abandoned their traditional fortes to try new and more radical designs. The Stearman Aircraft Company had created an impressive record with rather conventional biplanes, but the attack competition induced them to submit the almost radical (in terms of engineering features) X-1 00, a 269-mph bomber. Begun before the Boeing Company acquired Stearman, the X-100 was the first all-metal, twin-engine, retractable-gear airplane, powered by two experimental Pratt and Whitney R-2180 engines, Few firms were able to make such a severe transition, and the X-100 was an example. Subsequently, the new bomber was modified to the XA-21 and found to have very little better performance than the rejected XA-14. No other aircraft used the R-2180 engines, making the Stearman effort a true aerial dead end.

Grumman XP-50

As the money began to flow from Congress in 1939, there was an explosion of experimentation. Grumman had developed the squatty-nosed XF5F-1 for the Navy and a rather more handsome derivative, the XP-50, for Air Force use. The XP-50 was a hot airplane for the time, with a projected 424-mph top speed developed from two turbosupercharged Wright R-1820 engines. The aircraft was ill-starred, suffering first from a landing accident and then having the turbo supercharger blow up. The airplane augured in, and Air Force interest waned. The basic design, however, was developed into the XP65–canceled by the Air Force–and the XF7-F Tigercat, which went on to a great career with the Navy. Thus the effort of development was worthwhile, even though it didn’t lead to an Air Force purchase.

Beech XA-38 Grizzly

Some airplanes just look right, and the Beech Grizzly was one of these. Far more than a C-45 with a pituitary problem, the big XA-38 was powered by two R-3350 engines, giving it a 376-mph top speed, packing a 75-mm cannon plus two pairs of .50-caliber guns. The airplane had excellent flying qualities and almost certainly would have been ordered into production except for one problem: the R-3350 engines were required for the B-29 program, which, of course, had priority.

Brewster XA-32

The problems in experimental aircraft were sometimes the result of problems inherent in the manufacturing company. Brewster had startled the world when its pudgy Buffalo beat out Grumman’s entry in the first Navy monoplane fighter competition. After that, it scarcely did anything right; and the XA-32, despite a sound layout, became a compendium of management-induced faults. A husky brute of an attack plane, the XA-32 was terribly overweight at almost 20,000 pounds. The drag induced by its rotund shape was amplified by careless detail design, which left it festooned with bumps and lumps. It was underpowered by the Pratt and Whitney R-2800. The real problem, however, was that the XA-32 suffered from the terminally bad Brewster management system. First flight was not until 22 May 1943, two years after the design was proposed; and almost every aspect of performance fell short of the specifications. The firm was in such management shambles that it drew the wrath of Congress and actually went out of the aircraft manufacturing business.

McDonnell XP-85

Perhaps the best excuse for the XP-85 was the fact that it was ordered in October 1945 to be carried in the bomb bay of a B-36. The idea was that the XP-85 would be launched when an enemy fighter attack was imminent. It would engage, shoot down the opposing fighters, and then be picked up for storage inside the bomb bay again. McDonnell came up with a tiny, ugly airplane; its 21’1″ wings folded to a diminutive 5’5″ and, despite all the obstacles, it flew fairly well. Its launch and recovery problems brought about its demise. Its basic idea was picked up later with the RF-84 Fighter Conveyor program, but Air Force planners were driven to the conclusion that a long-range penetration fighter was needed–a concept that persists to this day.

Republic XF-12 Rainbow

Alexander Kartveli qualified as a Cellini-class artist with the design of the beautifully streamlined XF-12. Originally envisioned as a transatlantic passenger plane for Pan American World Airways, it lost out in the commercial market because of its small forty-four-passenger capacity. It was converted into a flying photographic laboratory.

Kartveli had addressed the problem of streamlining with consumate skill, and the Rainbow had aesthetically appealing lines. It suffered the ordinary development problems, including an engine fire and subsequent crash, but the real reason for its demise was the availability of both B-29 and B-50 types for interim duty as reconnaissance planes until the far more capable RB-47 was brought into service. The Rainbow is a perfect example of the importance of timing; had it been available in 1944, it almost inevitably would have been ordered in quantity, and the whole postwar structure of aircraft markets might have been altered, with Republic building follow-on airliners. As it was, the Rainbow disappeared into oblivion, despite its graceful lines and high performance.

Convair XP-81

An almost certain recipe for failure in aviation is the bet-hedger; anytime compromise is built into concept, success is almost impossible. Such was the case with the Convair XP-81, designed to have a GE J33 jet engine in the tail and a GE XT-31 turboprop in the nose. The idea, of course, was to combine the range of the turboprop with the dash speeds of a jet. In this case, doubling up the power plants didn’t double the pleasure but instead doubled the mechanical difficulties; and the outstanding promise of the P-80 and other pure jets relegated it to the drawing boards.

Convair XB-60

As development costs go up, so do the manufacturer’s fears–a combination that sometimes results in wistfully hopeful compromises like the XB-60. Ordered by the Air Force as a precaution against some unforeseen catastrophic failure in the B-52, the XB-60 had a 72 percent commonality with the veteran B-36. Essentially, a swept wing and tail were substituted, and eight J57 jet engines were added, resulting in a 508-mph bomber that was totally outclassed by the B-52. Convair bounced back, however, and proceeded with the beautiful supersonic B-58 Hustler, which carried bomber performance into a realm still not surpassed.

Douglas XB-43

The first American jet bomber was a straightforward derivative of the radical XB-42 Mixmaster, a twin-piston engine pusher aircraft of Learfan configuration that was supposed to do the job of the B-29 at about half the cost by achieving a 30 percent improvement in aerodynamic cleanliness. The jet age caught up with the XB-42, and Douglas responded by working the static test article into the XB-43 by substituting two GE J35 engines for the piston Allison V-1710s. The airplane was satisfactory, with a 51 5-mph top speed, not bad for a static article, but the B-45 was in the wings. The XB-43 subsequently became an engine test bed at Edwards, earning the nickname Versatile.

McDonnell XP-67

World War II changed American aviation from virtually a cottage industry to the greatest industrial effort ever seen, and a prime beneficiary of this process was the McDonnell Aircraft Corporation. Founded in 1939 with a total capitalization of $195,000, McDonnell built subassemblies for other manufacturers for most of the war. Its first military design, the XP-67, was a radical concept of blending wing and fuselage in a constant airfoil design. The airplane was perhaps the most beautiful–or sinister-looking–piston-engine aircraft of the Second World War. McDonnell was forced to use the brand new and unreliable Continental Xl-1430 engines, and these ultimately destroyed the aircraft and the program.

McDonnell had aimed high, with a pressurized cabin and armament provisions for either six 37-mm cannons or a single 75-mm installation. Top speed was guaranteed to be 472 mph, but the prototype achieved only 405 mph. The aircraft needed far more development time than the war allowed. Another critical factor was the rapid development of the P-51 Mustang, which was in mass production and seemingly able to handle any task assigned to it. As a result, the XP-67 was dropped and thus could be regarded a failure.

Nothing could be further from the truth, for the rapport McDonnell had established with both the Air Force and the Navy permitted the company to win a contract for a Navy jet fighter, the XFD-1 Phantom, starting a series that led ultimately to the giant McDonnell Douglas Company, with its enormous stable of civil and military aircraft.

THE costs of unsuccessful experimental aircraft are often cited as an example of waste or poor planning. In fact, the true value of an experimental aircraft may not be in the airplane itself, but in the team that it brings together for further efforts, for the concepts it proves incorrect, or in the spur that it gives to competition. Today the cost of experimentation has become perilously high, and there are fewer and fewer opportunities for companies to explore new but not yet proven lines. The truth of this situation is perhaps validated in a way that has never been seen in history before: one giant power, the Soviet Union, is apparently allowing its research to lag far enough behind the United States to permit certain developmental lines to prove themselves. It then steps in and builds, in large quantities, very similar aircraft types. In effect, our research and development is subsidizing Soviet R&D in a considerable way, and our experimental aircraft are, indirectly, also theirs.

National Air and Space Museum
Washington, D. C.


Walter J. Boyne (B.S., University of California, Berkeley; M.B.A., University of Pittsburgh) is Director of the National Air and Space Museum, Smithsonian Institution, and has held several previous positions on the museum staff. He is a retired USAF colonel, the author of several books and numberous magazine articles, and a previous contributor to the Review.


The conclusions and opinions expressed in this document are those of the author cultivated in the freedom of expression, academic environment of Air University. They do not reflect the official position of the U.S. Government, Department of Defense, the United States Air Force or the Air University.

Air & Space Power Magazine