flying wings 1870 to 1920
flying wings in Europe
tailless aircraft in the USA
German flying wings
British flying wings
Japanese flying wings
American flying wings in WW2
the first Northrop flying wing
Northrop flying wings in WW2
Northrop: towards the bombers
Northrop flying wing bombers
flying wings 1950s and beyond
secret Soviet flying wings

N-1M: the first Northrop flying wing
by E.T. Wooldridge

Jack Northrop became interested in the development of the cleanest possible airplane early in his career as an aircraft designer. In 1923, as an engineer for Donald Douglas in Santa Monica, California, Northrop continually explored advanced designs for aircraft, seeking new ways to eliminate drag and the severe penalties in aircraft performance it imposed. Even then, he envisioned an airplane without protruding surfaces that did not contribute in some way to lift. He even undertook the design of a tailless, all-wing glider as a "pastime" project, but never finished the aircraft, due to other commitments and lack of funds. In 1927, Northrop designed the incomparable Lockheed Vega as the best possible compromise that could be made with known and proven elements. Even this aircraft, with its conventional arrangement of wing fuselage and tail, gave hints of the unconventional concepts that were beginning to form in Jack Northrop's mind.

Northrop left Lockheed in 1928, and formed a small company, the Avion Corporation, in the Burbank/Glendale area, to further explore the idea of a tailless craft. His 1929 Flying Wing evolved, an aircraft unusual in appearance and performance, but more noted for its unique all-metal, stressed skin, multi-cellular construction. Although financial considerations forced suspension of further development of the airplane, its unique structure paved the way for major Northrop contributions to aviation in the perfection of all-metal construction.

Noted f or its multi-cellular construction, Northrop's 1929 Flying wing, with its twin booms and tail structure, was a cautious step toward his first true flying wing design in 1939.

In 1932, Northrop formed a new Northrop Corporation at El Segundo, California, in partnership with Douglas Aircraft. As Northrop continued to design and produce airplanes of a conventional nature, he found another opportunity to test his ideas with a wind tunnel model in 1937. He was assisted in his project, designated Model 25, by Edward H. Heinemann, who in his own right would have a profound impact on the design of military aircraft in the United States. Without any significant financial support, the project was abandoned after wind tunnel tests showed that it needed a tail.

In 1938, the Northrop Corporation became the El Segundo division of Douglas, with emphasis on production design. More interested in experimental design, Jack Northrop resigned, and in 1939 formed his own company once again, Northrop Aircraft Inc. Business came in the form of contracts for construction of Consolidated PBY subassemblies, a Norwegian order for 24 Northrop-designed N-3PB patrol bombers in March 1940, followed by a contract for co production of Vultee-designed Vengeance dive bombers. Northrop finally had the financial wherewithal, facilities, and ultimately, government interest, to enable him to pursue his interest in research and development and more specifically in the flying wing. As Northrop progressed through the early design stages of his first true flying wing, he sought the advice and technical expertise of one of the world's leading aerodynamicists, Dr. Theodore von Karman, Director of the Daniel Guggenheim School of Aeronautics at the California Institute of Technology (GALCIT), and von Karman's assistant, Dr. William R. Sears. Also available were all of the technical data and information in foreign aeronautical literature and NACA reports. Northrop and his assistant chief of design, Walter J. Corny, conducted extensive wind tunnel tests with a number of flying wing models. The result was an aircraft incorporating the latest thinking on buried engine design, new airfoil sections of low drag and improved stability, and the use of various high-lift devices, spoilers, and flaps.

Northrop viewed the diminutive N-1M (Northrop Model 1 Mock-up), with its 38-foot wing span, as a flying scale mock-up, perhaps one-third or one-half full-scale size. To achieve the efficiency and economy possible with a pure all-wing design, Northrop envisioned a commercial cargo airplane of a minimum 70-foot span, providing a maximum thickness of about 6 feet. With wing thickness growing in proportion to increase in span, a thickness of 15 to

20 feet would ultimately be possible. Depending on whether the use of the aircraft would be military or commercial, gun turrets, passenger cabins, and other special loads could be accommodated by bumps or projections that would not alter the basic characteristics of the aircraft.

The National Air and Space Museum is fortunate to have in its collection one of the original models built by Jack Northrop to investigate his theories of flying wing design. Constructed of balsa wood, tissue paper, and cardboard, the delicate structure bears a strong resemblance to the N-1M and incorporates many of the control surfaces evident in the real aircraft. Hinged wing tips, cardboard flaps that serve as rudders, and the elevons provide control and balance. The model was tested in hand-launched free glides to test its stability and flight characteristics.

The N-1M that evolved from many design studies and model tests was the first such tailless configuration to appear in the United States. The experimental aircraft was distinguished by the absence of any of the unusual appendages; the pronounced anhedral, or downward droop, of the wing tips gave the airplane a distinctly bird-like appearance. Aircraft configuration could be varied on the ground between tests to permit in-flight evaluation of the many variables associated with wing sweep, dihedral, and the all-wing design. In effect, the N-1M was the forerunner of today's "variable geometry" airplanes.

Control of the N-1M was accomplished using many of the same techniques and methods employed by the Hortens in Germany and other European designers. Elevons operated together for pitch control and differentially for roll control. Rudder control was accomplished initially with a plain split flap or "clamshell" at each wing tip. Actuated independently by the rudder pedals, they opened to produce drag, which, in turn, induced yaw. Both split flaps could also be opened simultaneously to increase gliding angle or reduce airspeed, thus serving in the role of air brakes.

The ICI-1M was of wooden construction, and thus easily adaptable to the many changes in configuration to which it was subjected during the flight test program. The aircraft was initially powered by two submerged 65-hp Lycoming 0-145 four-cylinder, horizontally-opposed engines driving two bladed pusher propellers by means of extension shafts. The engines, which were later replaced by 117-hp six-cylinder, air-cooled Franklin engines driving three-bladed propellers, were cooled by means of slot-type intakes in the leading edge of the wing.

Engineering and construction of the N-1 M took exactly one year, beginning in July 1939. The first flight of the N-1M, nicknamed the "Jeep," was in July 1940, and indeed was an accidental one, as pilot Vance Breese bounced the airplane into the air during a high-speed taxi run on Baker Dry Lake, California.

It took only several days of abbreviated test flights to prompt Jack Northrop to report encouraging results to Gen. H.H. "Hap" Arnold, Chief of the Air Corps. Northrop reported the airplane was both statically and dynamically stable about all three axes, with normal stick forces and good controllability, laterally and longitudinally. The aircraft was considerably shy of adequate rudder control using the trailing edge split flap rudders, but Northrop was optimistic about a solution.

All of the N-1 M's original distinctive design features are evident in these views of the aircraft at Muroc Lake in June 1941. Droop of 'the wing tips, with split flap rudders installed, could be adjusted on the ground; sweepback, dihedral, centre of gravity location (by changing angle of sweepback), and control surface arrangement were also adjustable. The faired tailwheel, which prevented excessive rollback on the ground, also provided some measure of directional stability. The metal bump on the top of the canopy was added to accommodate the pilot's head.

The N-1 M's test program provided valuable data for its successor, the N-9M, but it was not without problems. Choosing suitable power plants became the first, and most enduring dilemma, one that would plague many of Jack Northrop's piston-engined tailless designs.

Jack Northrop poses by the N-1 M, while Moye Stephens beams from the cockpit.

The flat surface of Muroc Dry Lake was the ideal site for early test flights of the "Jeep," shown here cruising at an altitude of 10 feet.

At an early stage in the test program, it was determined that the Lycoming engines were totally inadequate for the N-1 M. Moye Stephens, Northrop's company secretary and test pilot who took over the flight test program from Vance Breese, recalls that the Lycoming engines could not get the 4000-pound airplane any higher than ground effect during flights at the dry lake:

"In the initial flights with the Lycoming engines the ship would climb to about five feet and the increased induced drag associated with attempts to force it higher would bring it down to a landing. Continuous flight called for maintenance of a precise angle of attack. Any increase in the angle of attack and the ship would land. Any decrease in the angle of attack and the ship would land. The situation was complicated by a "dead area" in elevator effectiveness. In order to nose down it was necessary to move the wheel forward a disturbing amount with no response, and then the elevons would suddenly take over. In order to keep from banging into the ground it was then necessary to traverse the dead elevator area in the opposite direction to find the start of effectiveness. This was moderately unsettling while flying along five feet off the ground. I temporarily overcame the difficulty by use of the longitudinal trim flap: a control surface spanning the trailing edge of the centre section. With this adjusted to create a nose heavy condition, flight was maintained with a constant back pressure on the wheel. To nose down it was simply necessary to ease off the back pressure."

Dr. Theodore von Karman quickly came up with a solution for the problem. Realizing that the extremely thick wing was creating an airflow separation that was not coming together until aft of the wing, he suggested extending the trailing edge of the elevons into the closure of the airflow. The solution apparently had the desired effect.

Early test flights were made in a straight line over the length of the dry lake, generally at the maximum ceiling of the aircraft, about ten feet. On one of these flights, Stephens lost one foot of a propeller tip on the desert floor. Despite the extreme vibration which broke a rear spar, Stephens landed the ICI-1 M without further damage.

Replacing the Lycomings with 117-hp Franklins almost doubled the horsepower. The engines still had to be operated considerably in excess of the manufacturer's limitations to achieve anything approaching satisfactory flight. Overheating became chronic, and much time was lost in attempting to reduce oil and cylinder head operating temperatures to acceptable limits. By May 1941, Jack Northrop had come to the conclusion that the difficulty lay in engine design, since tests had shown sufficient pressure drop across the engines to cool them if they had been properly finned. Northrop considered the necessity of eventually changing engines once again, possibly using a new Lycoming six-cylinder engine of 150 hp. Apparently the switch never took place, since the N-1M was received by the National Air and Space Museum in 1950 with Franklin engines still installed.

Engine problems notwithstanding, Northrop concluded that the flying wing as demonstrated by the N-1 M was a practical idea, and entirely normal operation of the Flying Wing was no longer a problem. Although Northrop leaned toward a medium-range airplane as the next logical step, officials of the Army Air Corps (which officially changed its name to Army Air Forces on June 29, 1941) were turning to the possibility of a long-range airplane based on the flying wing principle. Requirements for an airplane with a range of 10,000 miles, cruising speed of 300 mph, service ceiling of 40,000 feet, and a bomb load of 10,000 pounds were soon being discussed in the context of a flying wing design. Northrop's feasibility studies eventually led to a conference with Air Force Materiel Division representatives in September 1941 to consider an experimental airplane with the desired military characteristics. The design that evolved was an incredible 140,000-pound behemoth; a far cry indeed from the tiny 4000-pound N-1M, which only two short years before had still been on the drawing board! Equally ambitious was an anticipated delivery date 24 months from contract approval.

As plans materialised for the long-range, heavy bomber, and as ICI-1M flight testing proceeded apace under tight security conditions in the desert, an ironic sequence of events occurred, hardly more than a footnote in the story of Jack Northrop's flying wings. For the Horten brothers in Germane, however, these events were a turning point in their own personal struggle to prove the flying wing concept.

Although the Air Force decided to classify all information pertaining to the Flying Wing in June 1941, routine publication of patent drawings had already occurred in the Official Gazette of the U.S. Patent Office on Ma \ 13. After media speculation forced an official release of information by Northrop and the Air Force, the patent drawings and an N-1M photograph eventually appeared in the international aeronautical journal Interavia on November 18,1941.

When the Horten brothers were interrogated after their capture by Allied forces in May 1945, they referred to the appearance of the N-1 NI photograph and drawings in Interavia as a stroke of good fortune. They used the article to "sell" the German Aviation Ministry on a more intensive program of development of the flying wing as a military aircraft in anticipation of American progress along these lines. The end result was the world's first turbojet-powered flying wing, the Horten Ho IX (Go 229), which flew in January 1945.

The drooped wing tips were eventually straightened on the N-1 M, as these two views of the aircraft illustrate. With maximum tip deflection and sweepback, and the N-1 M at rest on the tricycle landing gear, the wing tips were only one foot off the ground. Raising the nose during takeoff lowered the wing tips until they touched the ground while the tail wheel was still six inches in the air. Skinned wing tips during landing or takeoff occasionally resulted.

By November 1941, according to Jack Northrop, 200 flights had been flown with the N-1 M, which varied in time and altitude from a few seconds and 2 or 3 feet, to more than an hour and 7500 feet. During this time, Moye Stephens flew the N-1M with numerous combinations of wing tip deflection, dihedral, and sweepback. Initially it was thought that having the wing tips deflected downward would contribute to directional stability. It was soon found that they had little if any effect in this regard, but did lessen lift noticeably. Consequently, they were straightened.

With an aircraft of such radical design, stability was one of the primary concerns. In some configurations tested, yawing the aircraft, that is, moving the rudder pedals to cause motion about the vertical axis of the aircraft, and then releasing the controls, induced an oscillation called "Dutch roll." The oscillations were intense enough that Move Stephens wondered if they would reach a point beyond which they would be impossible to stop.

Eventually, with an alternative configuration of the aircraft, induced "Dutch roll" damped out after three or four oscillations.'

Of equal importance was the longitudinal stability of the flying wing. After much adjusting, experimentation, and flight testing, a configuration was found in which longitudinal, directional, and lateral stability were acceptable: straight wing tips, minimum dihedral, and the greatest degree of sweepback.

Throughout the test program, investigation of the controllability and stability of the N-1M was frequently hampered by poor performance and engine problems. The aircraft was overweight and underpowered, factors that required unusual approaches to some of the everyday problems associated with flight testing.