Cessna 172 N79438

The flight instructor and the light sport pilot, who was receiving training toward a private pilot certificate, were on an instructional flight in the airplane. Witnesses reported that, during takeoff, the airplane became airborne about 3/4 of the way down the 2,853-ft runway, and, when it was about 100 ft above the ground, the witnesses heard the engine lose power and subsequently regain it. One witness then heard the engine run "rough" and saw the airplane bank to the right and descend behind a treeline and out of view. Examination of the accident site indicated that the airplane impacted the side of a rail car parked on tracks beyond the end of the runway. The propeller blades displayed some leading edge damage and chordwise gouging, consistent with the engine producing partial power at the time of impact. No preexisting mechanical anomalies were found that would have precluded normal airplane or engine operation. The airplane's fuel selector was found in the "both" position; however, it had been moved by a first responder who could not recall its original position. Both wing fuel tanks contained fuel that appeared to be 100 low lead aviation fuel and was absent of water and debris. The gascolator was destroyed, but the damaged carburetor contained a small amount of fuel, which was tested and found not to have water in it. The carburetor heat control was found in the "ON" position. Fuel starvation due to an improperly positioned fuel selector could have resulted in the loss of engine power, and testing was conducted to determine if this was a possibility. There were no onboard recording devices, but airport security cameras captured segments of the airplane's start, taxi and takeoff roll. Using a similar airplane, the fuel selector was moved to the off position at various power settings, and the time it took for the engine to lose power was measured. Comparing these measurements to the times obtained from the security camera recordings indicated that the accident airplane's fuel selector could not have been turned to and remained in the off position during engine start or run-up. The timing comparisons did, however, indicate a possibility that the fuel selector could have been moved to the off position just before takeoff, which could have resulted in the loss of engine power. If this was the cause of the power loss, changing the position of the fuel selector in flight may have allowed the engine to regain at least partial power before impact. Diminished airflow through the carburetor due to carburetor ice also could have resulted in the loss of engine power. Temperature and dew point values were conducive to carburetor icing at taxi power settings. If carburetor ice formed during the airplane's taxi with the carburetor heat off, the additional engine heat from the application of takeoff power could have caused the ice to break off and block air flow during the airplane's initial climb, resulting in the loss of engine power. The subsequent addition of carburetor heat could then have melted some or all of the remaining ice and allowed the engine to regain at least partial power before impact. Although no preexisting mechanical anomalies could be found, the amount of damage to the airplane and the lack of recording devices precluded a definitive determination of how the engine power was lost and at least partially recovered.  

Quoted verbatim from the NTSB record.