N808DDBobbi Boucher DUO DEUCE · 2016

On June 6, 2016, at 1759 eastern daylight time, an experimental, amateur-built Duo Deuce, N808DD, was substantially damaged during a collision with terrain after takeoff from Stafford Regional Airport (RMN), Stafford, Virginia. The commercial pilot, who was also the owner/builder was seriously injured. Visual meteorological conditions prevailed, and no flight plan was filed for the initial test flight, which was conducted under the provisions of 14 Code of Federal Regulations Part 91.In an interview with a police officer just after the accident, the pilot reported that during the initial climb after takeoff, both engines experienced a "sudden" loss of power. She identified an open area for a forced landing, and upon touchdown, the airplane nosed over and came to rest inverted.

In an interview with a Federal Aviation Administration (FAA) aviation safety inspector, the pilot stated that the engine run-up was "normal and everything was fine." During the takeoff roll, the airplane reached "flying speed," the nose lifted from the runway, and she advanced the throttles to 2,700 rpm. As the airplane became airborne, the "left wing came up but the right wing was down." The pilot stated she added left rudder and left aileron, but realized the airplane was "eating up a lot of runway" and there was insufficient runway remaining on which to land. She turned the airplane right to avoid the interstate highway that ran perpendicular to the runway beyond the departure end. According to the pilot, "I stalled I guess, I hit the ground pretty hard."

In a subsequent telephone interview with an NTSB investigator, the pilot stated she did not recall the conversation with the police officer, and stated that the right engine stopped producing power. In a media interview several months later, the pilot reported she remembered "every detail" of the accident flight, and said that when she taxied the airplane onto the runway at RMN, the purpose was to perform a "high speed taxi" and that the subsequent takeoff was "inadvertent." After takeoff, the airplane experienced "engine failure" followed by an aerodynamic stall.

Several witnesses provided written statements. One witness was well-acquainted with the pilot and said that he was there to assist her with the flight. In his statement, he referred to the flight as both a "test flight" as well as the "first flight" for the airplane. The witnesses described the takeoff and climb as "slow," stating that the airplane was "wobbling" and the wings were "rocking." One witness estimated that the airplane climbed to about 300 feet above the runway before it slowly descended.

A review of videos recorded from two airport security cameras, as well as an on-board video recorded with the pilot's cellphone revealed a shallow takeoff and initial climb. Almost immediately after takeoff, the airplane's track diverged from the runway centerline off the right side of the runway and over the grass apron. The climb stopped at what appeared to be treetop height, the wings rocked, and the airplane continued to pitch up as it descended until ground contact. The instrument panel could not be viewed, but the propeller speeds appeared constant and both propellers appeared to be turning at the same speed during the takeoff roll and the entire flight until ground contact.

Examination of photographs revealed the airplane remained largely intact, with the left engine separated. Both wings and the tail section were substantially damaged.

According to FAA records, the pilot held a commercial pilot certificate with ratings for airplane single and multiengine land and sea, and a flight instructor certificate with ratings for airplane single and multiengine. She was issued an FAA second-class medical certificate on October 29, 2015, and reported 6,420 total hours of flight experience on that date.

The two-seat, twin-engine, low-wing airplane, equipped with two Lycoming IO-320-B1A engines was initially registered in 2013, and issued an airworthiness certificate on March 19, 2016. The pilot modified a Van's Aircraft, Inc., RV-8 single engine airplane kit. Instead of the nose-mounted, single-engine configuration for which the kit was designed, the airplane was configured with two wing-mounted engines.

Examination of maintenance records revealed the engines were previously owned, and installed on an airplane that was involved in an accident on February 19, 2008. Each engine experienced a propeller strike event during that accident sequence. The records did not indicate that a mandatory sudden-stoppage inspection had been performed on either engine after the event and prior to their installation on N808DD.

The airplane's instrument panel was equipped with an iPad mount, and a telephone was mounted above and behind the pilot. Both the iPad and the telephone were requested so that the original media could be examined. The pilot refused to provide either device; however, she provided a 33-second-long video file, which was consistent with the vantage point of the cockpit mounted cell phone.

Throughout the takeoff roll, flight, and subsequent impact with terrain, the effect of the video camera's rolling shutter effects on the representation of each propeller did not substantially change. Because the distortion of each propeller due to rolling shutter was consistent, the recording suggested each propeller's rpm remained at an unquantified but mostly steady state.

The airplane was subsequently recovered to Shannon Regional Airport, Fredericksburg, Virginia, where it was examined by representatives of the NTSB and Lycoming Engines. In addition, a test run of the right engine, which remained mounted in its nacelle and attached to the airframe was performed.

The constant-speed propeller was damaged during the accident sequence. It was removed, an expansion plug was seated in the front of the crankshaft, and a fixed-pitch propeller was mounted. The airplane was pushed out to the taxiway apron, jumper cables were attached to an airport service vehicle and the airplane's battery, and an engine start was attempted utilizing the airplane's own fuel system.

The engine was started, and it ran smoothly and continuously until engine oil sprayed in the propeller wash. The engine was stopped, the propeller was removed, the crankshaft expansion plug was reseated, the engine was serviced with oil, and another engine start was initiated.

The engine started and idled smoothly, and ran continuously without interruption. The throttle was increased and decreased, and the engine ran smoothly through the power changes. Engine oil pressure and fuel flow indications were consistent with the throttle position as it was changed.

With full throttle application, approximately 2,200 rpm was observed on the aircraft tachometer. The engine ran roughly, consistent with a lean fuel/air mixture setting. Fuel flow as noted on the digital flow meter was approximately 19 gallons per hour (gph). Typical fuel flow requirements for the subject engine operating at this power setting would be 8.5 gph based on a lean limit mixture setting and approximately 10 gph based on a best power mixture setting. The abnormality noted with the digital flow indication system was consistent with air entering the fuel system, resulting in acceleration of the flow scan vein and high fuel flow indications. The fuel flow abnormality as noted was consistent with air entering upstream of the engine on the airframe side of the fuel system.