Primary finding
Probable cause
Slippage in the engine clutch and gear reduction module and a resultant uncommanded movement of the propeller blades into reverse thrust.
Investigator assessment
Analysis narrative
Family members and friends reported that, 2 years before the accident, the experimental, amateur-built airplane had experienced a serious in-flight engine problem. Since that time, the pilot had been repairing the engine and propeller system, and the accident flight was the airplane's first flight following the reported engine problem. The pilot-rated passenger reported that, just after departing for a short local flight and turning the airplane to enter the downwind leg of the airport traffic pattern, the pilot informed him that there was a problem and that they needed to return to the airport. He added that the pilot began working "frantically on an electric switch for the propeller" while the engine was "screaming" but that the airplane continued descending. The airplane struck a tree and then impacted a house. An explosion and a postimpact fire immediately ensued, which consumed most of the airplane and the house. The pilot and passenger were able to exit the burning wreckage without assistance. An examination of the engine clutch and gear reduction module found some metal contamination and a nonstandard displacement of the sprag clutch elements, and these issues likely resulted in clutch slippage, which can lead to the uncommanded movement of the propeller blades. The propeller blades were observed to be in about 5 degrees of reverse pitch, which would result in a substantial loss of glide capability.
Source record
Factual narrative
At 0925 N513AC made a radio transmission requesting to depart BAK and remain in the traffic pattern for touch-and-go landings. At 0932 the tower controller at BAK cleared the N513AC airplane for takeoff from runway 23 and N513AC responded. There were no further radio transmissions received from N513AC. About 0933 other aircraft in the traffic pattern radioed that an airplane had crashed about 1 mile south of BAK and was on fire. The aircraft rescue and fire fighting (ARFF) crew at BAK was monitoring those radio transmissions and they immediately responded to the accident scene. The FAA Airport/ Facility Directory, East Central U. S., indicated that BAK was a towered airport with a field elevation of 656 feet mean sea level (msl). Runway 05-23 was a 6,401 feet long by 150 feet wide concrete runway. Runway 14-32 was a 5,001 feet long by 100 feet wide asphalt runway. The propeller, propeller governor, and clutch and gear reduction module were examined at an FAA certified repair station in Sellersburg, Indiana on August 13, 2013. The electrically operated propeller pitch change motor was thermally damaged and damage prevented an operational check. There was no digital propeller pitch gauge circuit board installed in the propeller extension recess; and there was no "third wire" to direct the signal to the innermost commutator slip ring. On disassembly it was noted that there was no reverse pitch lockout spacer installed, as recommended by the propeller manufacturer in applications when reverse pitch operation is not desired. During examination the propeller was observed to be in the Beta (reverse) pitch position and was estimated to be at about 5 degrees of reverse pitch. The NSI A30 Module output shaft was removed for examination and noticeable differences were observed between the unit being examined and the manufacturer's drawings. On the one piece shaft with integral propeller hub, spanner nuts were used on the aft portion of shaft instead of the bolt and nut fastener called for in the drawings. The thermally damaged clutch appeared to grab in the correct direction, however some metal contamination was found in the clutch. During disassembly, the shaft and bearing areas were examined and a nonstandard displacement of the linear cam device (LCD) sprag clutch elements was noted. Original assembly specifications and other documents from NSI Aero Products, the propeller manufacturer, showed that the electronic control unit took inputs from the throttle position, rotation rate, and manifold pressure information to calculate the optimum propeller position. It would then process this information combined with manual inputs to propeller's electric motor driver to activate the propeller blade angle from about 29 degrees in "coarse" pitch to about 10 degrees in "fine" pitch. The propeller could also be controlled to a "reverse" pitch of about 25 degrees. Those documents showed that the propeller would be locked out of the Beta position, or reverse pitch position, unless the propeller RPM dropped below about 1,300 RPMs. The documents showed that, if a mechanical reverse pitch lockout mechanism was not installed, and if the RPM dropped below about 1,300 RPM then the propeller could go into Beta pitch or reverse pitch, even when in flight. FAA Advisory Circular AC 35-1, Certification of Propellers, Section 1.2, defines Beta control as "a system whereby the propeller blade angle is directly selected … (and) is normally used during ground handling, including reverse pitch angles". It also defines Reverse pitch as "any blade angle below ground idle blade angle", and defines a Reversible pitch propeller as "a propeller in which blades can be rotated to a reverse pitch blade angle while operating". An autopsy was not performed on the pilot. Specimens from the pilot were not available and forensic toxicology was not performed. At 0938 the Automated Surface Observation System at BAK reported calm wind, visibility 10 miles, clear of clouds, temperature 17 degrees Celsius (C), dew point 14 degrees C, and an altimeter setting of 30.09 inches of Mercury. Evidence at the accident scene showed the airplane was upright and generally eastbound when it struck a tree about 25 feet above ground level (agl), which was the initial impact location. Debris from the tree led to the east about 40 feet where at about 15 feet agl the airplane impacted on the roof of the south side of the single story house. The airplane continued moving through the roof and ceiling and partially exited through the east wall of the house with the aft portion of the airplane still inside the damaged house. The airplane came to rest upright and slightly left side down with the nose oriented to the east. The total distance from first impact on the tree to the final resting location was about 60 feet. The airplane was almost completely consumed by fire and several burned components of the house were on top of the airplane wreckage, with many other burned or partially burned items from the house scattered around the back yard. Fractured portions of both all-composite propeller blades were found at the scene and east of the tree. The damaged propeller tips and blade portions showed slight evidence of leading edge impact damage and some evidence of chord-wise smearing on the blade face. Both propeller blade tips were separated about 4 inches from the tip. One propeller blade was impact separated at the blade root near the propeller blade grip. The other blade also broken near the propeller blade grip, but remained attached and was bent aft about 30 degrees. The propeller spinner displayed impact damage mostly on only one side with only slight evidence of rotational smearing. The engine to fuselage mounts did not exhibit significant impact damage, remained intact, and the engine remained attached to the mounts. The exhaust system remained attached and did not exhibit significant impact damage. The engine, gear reduction housing, propeller shaft, propeller hub, and propeller spinner remained attached. Mechanical continuity of the throttle cable was confirmed from the firewall to the throttle arm at the front end of the engine. All portions of the airplane were accounted for at the accident scene and all flight control surfaces and flap surfaces remained attached or partially attached. Damage to the wreckage prevented an examination of the ignition system or the induction system components and prevented a complete assessment of flight control continuity. The high-wing, fixed tricycle landing gear, single-engine airplane, serial number (s/n) 5330, was completed by the pilot as a kit-built airplane. In 2006, it was issued an FAA airworthiness certificate in the experimental category as an amateur built airplane. The airplane was powered by a 135 horsepower Subaru EJ22 automotive engine, s/n 642126. The fuel injected engine was mated to an NSI A30 gear reduction module, which drove an electrically operated NSI CAP model 200 2-blade composite controllable pitch propeller. The airplane was constructed with a welded steel fuselage frame with a fiberglass covering. It had aluminum wings of conventional design, and had two cockpit seats in a side-by-side configuration. An airplane weight and balance document dated August 31, 2006, showed the airplane then had an empty weight of 1,292 pounds, a maximum gross weight of 1,960 pounds, and a maximum fuel capacity of about 50 gallons. Family members and friends reported that two years before the accident the airplane had some sort of in-flight engine problem which resulted in a successful forced landing at Decatur Airport (DEC), Decatur, Illinois. On October 11, 2011, the airplane was then partially disassembled and trailered from DEC to the pilot's hangar at BAK. The pilot, who was also the original builder, had for the past two years been actively working to fix that engine problem. On July 25, 2013, at 0933 eastern daylight time,