Accident Overview
On October 26, 2023, at approximately 1330 Pacific daylight time, an experimental amateur-built Vans RV-12, registration N28BS, was substantially damaged during a forced landing on a road near Novato, California. The pilot and passenger were not injured. The flight was conducted under Title 14 Code of Federal Regulations Part 91 as a personal flight.
Pilot Report
According to the pilot, the accident flight was the first since maintenance that included a condition inspection, rebuilding of both carburetors, and replacement of ignition leads for cylinders Nos. 1 and 3. After an uneventful pre-takeoff engine run-up, takeoff was initiated on runway 31. The pilot noted that the airplane seemed "lethargic" during the takeoff roll. When the airplane reached about 100 ft above ground level, the engine began to run rough and lost partial power. The pilot initiated a 180° turn but found the airplane too high and too fast to land on runway 13, so he turned left toward a nearby road. The airplane landed hard on the road, causing the nosewheel landing gear to collapse.
Engine Data and Postaccident Examination
The airplane was equipped with a Dynon FlightDek D180 display that recorded engine parameters. Downloaded data showed that during the takeoff sequence, engine rpm decreased from 5,000 rpm to about 3,200 rpm, and exhaust gas temperatures for cylinders Nos. 1 and 3 dropped from about 1,350°F to between 350°F and 500°F. At the time of decreased rpm, fuel pressure was 2.3125 psi and fuel flow was 5.625 gallons per hour. According to the Vans RV-12 Pilot Operating Handbook, fuel pressure limits were 2.2 to 5.8 or 7.2 psi, depending on the engine-driven fuel pump installed.
Postaccident examination revealed structural damage to the fuselage. The aircraft had a nonstandard Aircraft Service in-flight mixture control system for the Rotax engine. The mixture control knob was found one turn out from full rich. The carburetor float chamber vent line for cylinders Nos. 1 and 3 was found disconnected. The mixture system jet was free of obstructions. Rotational continuity and thumb compression were verified on all cylinders. Compressed air applied to the fuel inlet lines showed no obstructions. Both carburetors were partially disassembled; slides moved freely and floats were intact.
Mechanic's Testimony and Run-up Test
The mechanic who rebuilt the carburetors stated that the mixture control system allowed leaning to obtain a 100-200 rpm rise with one or two turns of the knob, but he advocated using it only above 4,000 ft. He indicated that at two turns out, exhaust gas temperatures rise but remain within limits, and leaning beyond two turns causes roughness but not power loss. The mechanic performed a run-up test on his own similarly equipped airplane. He reported negligible effect of running the engine with the carburetor float chamber vent line disconnected for cylinders Nos. 1 and 3 during a simulated 5-minute climb. Operating the mixture control through its full travel had an insignificant effect; the engine felt and sounded normal throughout the test.