Accident Overview
On 27 February 2001, a Beech C23 (registration G-BBTX) was being used for a private flight at Blackbushe Airport, Hampshire. The aircraft had two people on board—a pilot and an instructor—and was conducting a single-engine piston class rating dual flight as part of the pilot's licence revalidation. The flight departed from Blackbushe and after refresher training in the local area, returned to the airfield. Weather conditions were described as good, with runway 08 in use, surface wind from 090° at 5 kt, and temperature/dew point of +3°C and +2°C respectively.
Sequence of Events
The first practice forced landing (PFL) from overhead the airfield at 2,000 ft was completed successfully. During the descent with idle power, carburettor heat was selected on and engine warming was performed several times. A second PFL from a similar position was then attempted. During final approach, it became evident the aircraft would not reach the runway. At approximately 100 ft above ground level, the instructor called for a go-around. The pilot was already applying power, but the engine did not respond. During the attempt to reach the runway, airspeed decayed, the aircraft stalled, and it landed heavily on the paved undershoot short of the displaced threshold after clearing an earth bank. Both pilots vacated the aircraft normally; no injuries were reported.
Aircraft Background
G-BBTX was owned by a flying group of ten members. In September 2000, the aircraft suffered propeller damage from a heavy landing, requiring engine removal for shock loading inspections. After reassembly with a new camshaft, tappet bodies, and bearings, the engine was test run and reinstalled. During that work, the carburettor heat flap box was replaced due to excessive play in a spindle bearing, and the heat flap return spring was also replaced (it had been broken). Several subsequent flights were made by group members; three of these experienced rough running during magneto checks and when leaning the mixture. The mixture control was found overly sensitive and was adjusted. One group member emailed others about these engine problems and logged them in the aircraft sheets, but the accident pilot stated he had not received any notification and did not recall seeing the logs. He said he would not have performed PFLs had he known.
Examination and Testing
After recovery, examination revealed the nose and left main landing gear collapsed, damage to the left wing and propeller. One propeller blade tip was bent forward; both tips showed chordwise scoring, indicating the engine was producing power on ground contact. Fuel was present and uncontaminated. Throttle, carburettor heat, and mixture controls operated satisfactorily. Ignition timing and mechanical fuel pump function were correct. No defects were found onsite.
The engine was removed and test-run at the same facility that had performed the earlier work. After confirming crankshaft flange runout within limits, the engine started and ran normally. Magnetos, acceleration checks, idle settings, and power output were all satisfactory—comparable to the previous post-repair test.
Discussion
No defects were identified that could explain the engine's failure to respond to throttle inputs on the second PFL. The symptoms were more consistent with carburettor icing. According to CAA guidance, the ambient conditions (temperature +3°C, dew point +2°C) were conducive to severe carburettor icing at any power setting. The aircraft was operated at low throttle settings during the PFLs, increasing susceptibility to icing. Additionally, at low power, exhaust manifolds cool quickly, reducing heat available for the carburettor heat system. Ironically, the earlier replacement of the carburettor heat flap box and broken return spring (which had allowed warm air to flow continuously) may have made the engine more prone to icing.
The AAIB noted that about six accidents annually may be attributable to carburettor icing, and the rate has remained constant. The aircraft was not fitted with a carburettor ice detection system.
Conclusions
No defects were found that could account for the engine not responding to throttle. The weather conditions indicated severe carburettor icing possible at any power setting. The most likely cause of the accident is believed to be carburettor icing.