N105JFCessna 208 · 1997

On December 21, 2014, about 1030 Pacific standard time, an experimental amateur-built light sport Farthing James W II (Spectrum Aircraft) Beaver RX-550, N105JF, struck the ground in a field following a partial loss of engine power shortly after takeoff from Perris Valley Airport, Perris, California. The airplane was registered to, and operated by, the pilot/builder as a personal flight under the provisions of 14 Code of Federal Regulations (CFR) Part 91. The private pilot sustained serious injuries, and the airplane sustained substantial damage to the forward fuselage and both wings during the accident sequence. Visual meteorological conditions prevailed, and no flight plan had been filed.

The airplane departed to the south from the Perris Valley Airport ultralight airstrip, located adjacent to the approach end of the airport's main runway (33). A witness, who was positioned at the hangars parallel to the airstrip, watched as the airplane climbed to about 100 feet above ground level (agl) and began a right crosswind turn. He stated that once the turn was completed the engine started to "surge," and then lost all power. The airplane began to descend, and when it was about 15 ft agl, the right wing dipped and struck the ground. The airplane then cartwheeled, and came to rest upright on its landing gear.

The pilot reported that during the 6-month period leading up to the accident, the engine had been emitting a "surging" and "pulsating" sound during takeoff. There was no discernable loss of power, and the pilot estimated that the engine would lose about 50 rpm during those events. During the 2-week period leading up to the accident, he modified the fuel line routing, and serviced the carburetors in an effort to resolve the problem, which he attributed to a lean fuel mixture condition.

The first flight following that maintenance was performed at 0830 on the morning of the accident. The takeoff was uneventful, and the engine performed normally without surging; the pilot flew in the traffic pattern operating the engine for 0.6 hours. The pilot reported that for the accident flight later that morning, the engine performed normally during takeoff, but as soon as he began the crosswind turn, the engine started to make a "hollow" sound, accompanied by a partial loss of power. He turned on the electric fuel pump, with no change in engine performance, as he attempted to maneuver the airplane through a shallow turn while avoiding telephone lines. The pilot stated that during the descent, he stalled the airplane about 20 ft agl, and it landed hard on its right side.

The high-wing two-seat airplane was constructed primarily out of aluminum and steel tubing, and covered with fabric. The two-cylinder, two-stroke Rotax 582 series engine was equipped with a composite propeller, and mounted behind the seats in a "pusher" configuration. The engine was not covered by a cowl.

The high-wing two-seat airplane was constructed primarily out of aluminum and steel tubing, and covered with fabric. The two-cylinder, two-stroke Rotax 582 series engine was equipped with a composite propeller, and mounted behind the seats in a "pusher" configuration. The engine was not covered by a cowl.

The airplane was examined by representatives of the NTSB, the Federal Aviation Administration, and the engine manufacturer Rotax. A complete examination report can be found in the NTSB public docket.

Both wing-mounted fuel tanks contained fuel, and fuel was present in all supply lines through to the carburetors. Internal inspection of the engine-driven fuel pump revealed that it was full of fuel, and that all internal seals and membranes were intact. The fuel selector valve was in the "OFF" position, however, according to family members of the pilot, it was switched off by his son shortly after the accident.

The front (magneto-side) carburetor float chamber contained only a residual quantity (about 1 tablespoon) of fuel. The rear float chamber was about 1/2 full. The fuel in the carburetor was checked for the presence of water utilizing SAR-GEL water-detecting paste; no water was detected.

No evidence of any catastrophic engine failures or seizures was observed during the examination; however, evidence of a series of installation and maintenance errors was noted.

The red sieve sleeves in both carburetor float chambers appeared to be crushed along their longitudinal axes, in a manner consistent with them not being secured correctly when the chambers were installed. According to the Rotax representative, the sieve screens' purpose was to separate foam/air bubbles out of the fuel prior to entering the carburetor. The float bracket inside the front carburetor hung about 1/32 inch lower than the rear carburetor's float bracket, indicating an adjustment imbalance between carburetors, and neither bracket was parallel to the float chamber as required by the Rotax Repair Manual.

Each carburetor was equipped with two vent duct openings at the float chamber. Rotax installation and maintenance manuals stated that the two ducts should not be directly vented to ambient pressure, but instead be connected to each other with a pressure compensating tube (or a tube containing small "compensation bores"). The manuals also stated:

"Equal air pressure conditions must prevail in the vicinity of the carburetor air intake and the carburetor float chamber vent pipes (i.e. the pressure must not be influenced by the propeller air stream). If necessary... the carburetor vent pipes should be routed to a calm air zone or connected to a vent chamber..... The vent duct openings to atmosphere are provided with a hose which prevents ingress of dust and water into the carburetor. Velocity pressure must never act upon venting ducts as it would influence pressure in float chamber, and therefore alter the air/fuel mixture."

Examination of the carburetor venting revealed that instead of utilizing the carburetor compensation tube, the carburetor's vent ducts were routed directly to ambient air with plastic tubing, which passed underneath the bowl spring clips, and laterally outwards into the air stream.

The only fuel filter installed was a plastic-screen type, which was positioned on the "suction" side of the fuel system, just after the selector valve, and before both the electrical and engine-driven fuel pumps. This installation was in conflict with the Rotax Installation Manual, which required the filter to be installed between the electrical fuel pump and the carburetors.

With regard to the fuel line, the manuals stated:

"This section of pipe is under considerable negative pressure when the engine is running hard, and the tiniest flaw in any joint will cause air to be sucked into the system, considerably reducing the capacity of the fuel pump. Air leaks are much more dangerous when the fuel tank is mounted below the fuel pump and carburetor."

Examination of the airplane revealed that the wing-mounted fuel tanks were positioned about 24 inches below the carburetors, and fuel staining was present at the fuel primer handle, indicative of a fuel leak in the fuel supply system.

Subsequent to the examination, the carburetors were reassembled, and the engine was configured for a ground run. The engine started on the second attempt, and was operated throughout its speed range. Multiple runs were performed, and during one run the engine temporarily "hesitated" while operating at high speed. On another run fuel was observed to flow out of the aft vent line of the forward magneto-side carburetor (the same carburetor which was found with an almost empty bowl following the accident). Fuel flowed from the aft vent line irrespective of whether the electrically-driven fuel pump was operating or not.