27 September 1946

Geoffrey Raoul de Havilland, Jr. (Sport & General Press Agency, Ltd, 1 September 1946; © National Portrait Gallery, London. NPG x184369)

27 September 1946: Geoffrey Raoul de Havilland, Jr., O.B.E., Chief Test Pilot of the de Havilland Aircraft Co., Ltd., and the son of the firm’s founder, was killed during a test flight of a prototype DH.108 Swallow, TG306.

Geoffrey de Havilland, Jr., in the cockpit of the second DH.108 Swallow prototype, TG/306. (Flight)
Geoffrey de Havilland, Jr., in the cockpit of the second DH.108 Swallow prototype, TG306. (FLIGHT)

De Havilland had taken off from the company airfield at Hatfield at 5:26 p.m. for a planned 45 minute flight. Flying over the Thames Estuary, east of London, England, de Havilland put the swept-wing jet into a high-speed dive from 10,000 feet (3,048 meters). As it approached 5,000 feet (1,524 meters) at 0.88 Mach, (658 miles per hour, 1,060 kilometers per hour), the shock waves building up along the wings’ leading edges disrupted the air flow over the wings, causing them to stall. TG306 pitched violently downward. A NASA report called this “. . . an undamped violently divergent longitudinal pitching oscillation at Mach 0.875. . . .”  The extreme aerodynamic loads cracked the main spar and both wings failed. The DH.108 crashed into Egypt Bay, Gravesend, Kent.

The wreck was located the following day. The body of Geoffrey de Havilland was found ten days later. He had suffered a broken neck and fractured skull as a result of his head striking the canopy during the violent oscillations of the aircraft.

(Grace’s Guide)

FLIGHT reported:

Geoffrey de Havilland was one of the outstanding test pilots in the country, and his work has played a vital part in the perfecting of such noteworthy types as the Mosquito, Hornet, Vampire and 108. His death is a serious blow not only to the company but to the country, for in the exploration of the unknown threshold of sonic flight, a combination of skill and cool courage are qualities demanding the utmost of test pilots. Geoffrey de Havilland had these qualities in a very high degree.

FLIGHT and AIRCRAFT ENGINEER, No.1971, Vol. 1, Thursday, 3 October 1946, at page 364

De Havilland DH.108 TG/306. (Unattributed)
De Havilland DH.108 TG306. (Unattributed)

The DH.108 was a single-seat, single-engine jet fighter prototype with swept wings and no conventional tail. It was similar in configuration to the Messerschmitt Me-163 rocket-powered interceptor. The first two prototypes, TG283 and TG306, were built using production English Electric DH.106 Vampire F.I fuselages. TG283 had a 43° sweep to the wings’ leading edge, while TG306 had a 45° sweep. The airplane was powered by a de Havilland Goblin 3 centrifugal-flow turbojet engine (a development of the Halford H.1) which produced 3,350 pounds of thrust (14.90 kilonewtons).

The first and third DH.108s also crashed. VW120 was destroyed on 15 February 1950 when it crashed after a dive. The left wing had separated and the pilot, Squadron Leader Stuart Muller-Rowland, also suffered a broken neck as a result of the airplane’s violent oscillations. On 1 May 1950, while conducting low-speed tests, TG283 went into an inverted spin. Squadron Leader George E.C. Genders, AFC, DFM, bailed out but his parachute did not open before he hit the ground and he was killed.

Geoffrey de Havilland, Jr., exits the cockpit of one of the company's jet aircraft. (Photograph Courtesy of Neil Corbett, Test and Research Pilots, Flight Test Engineers)
Geoffrey de Havilland, Jr., OBE, exits the cockpit of a DH.108 Swallow prototype. (Photograph courtesy of Neil Corbett, Test and Research Pilots, Flight Test Engineers)

© 2018, Bryan R. Swopes

Share this article:

About Bryan Swopes

Bryan R. Swopes grew up in Southern California in the 1950s–60s, near the center of America's aerospace industry. He has had a life-long interest in aviation and space flight. Bryan is a retired commercial helicopter pilot and flight instructor.

4 thoughts on “27 September 1946

  1. Test pilot Eric Brown flew the DH.108 and survived. In his autobiography “Wings On My Sleeve” (pages 195-196) he describes what it was like when the aircraft departed from normal flight:

    “At mach 0.88 it happened. The ride was smooth, then suddenly went all to pieces. As the plane porpoised wildly my chin hit my chest, jerked hard back, slammed forward again, repeated it over and over, flogged by the awful whipping of the plane. My thoughts were grim. This was how it happened. This was how he had died.”

    Brown attributed his survival to the fact that he was shorter than de Havilland, so his head hadn’t been thrown against the frame of the canopy with enough force to kill or incapacitate him. Such were the small details that could mean the difference between life and death in the pioneering age of jet aviation.

      1. Thanks to AndrewZ for describing Eric Brown’s experience of the DH108. Brown generally experienced dangerous flight characteristics with tail-less aircraft, describing them as “killers”. The Northrop wings were similarly marginal in stability and on the edge of critical divergence at quite low Mach numbers. Even the Me163 went into a nose-down “death tuck” at around Mach 0.80. You can see why perhaps the most successful “flying wing” expert, Alexander Lippisch, moved away from that concept to the delta wing, during the war years. That led to the XF92 post-war and the subsequent F102, F106 and B58. Really, he just “filled in the V” behind the trailing edge of a flying wing, and you can see how that improved longitudinal stability and reduced sensitivity to centre of pressure movement. Now we see the B2 and B21, people rightly point out that these are now possible because of modern electronic flight control systems, which provide constant, instantaneous correction of any divergence. Even so, they are not aerodynamically superior, the advantages are primarily in low radar profile and optimised “stealth” characteristics. De Havilland originally planned the 106 Comet as a tail-less design and the 108 was largely intended to test the principle. You can see why they changed their minds on that one! However, you can also see in the Sea Vixen particularly, almost a “flying wing” with a conventional tail attached through the twin booms. Proving the point that with a tail-less design, the best way to make it stable is to stick a tail on it.

Comments are closed.