ERA21LA080

The pilot’s failure to use carburetor heat in environmental conditions favorable for serious carburetor ice during a prolonged wait with the engine at a low power setting before takeoff, which resulted in a partial loss of engine power due to carburetor ice. Also causal was the pilot’s failure to recognize degraded engine performance during the extended takeoff roll and abort the takeoff.

The pilot was initiating a cross country flight with three passengers on board. After performing an engine run-up that included a functional check of the carburetor heat and verification of full static rpm, he waited 8 minutes for his instrument takeoff clearance with the engine operating about 1,200 rpm. After being cleared for takeoff, he taxied onto the runway, applied full throttle, and began his takeoff roll. One pilot-rated witness, who was also an airframe and powerplant mechanic, reported hearing a loud noise that he described as a “definite hard miss,” and a second pilot-rated witness reported hearing “popping and banging” throughout the airplane’s takeoff. The pilot reported attaining a normal takeoff distance, which would have been about 1,135 ft according to performance calculations; however, the second witness noted the airplane rotated about 1,542 ft down the 3,350-ft-long runway. After rotating, the pilot pitched for 73 knots, and at 100 ft, he reported the airplane would not climb. The airspeed started to decrease, which resulted in the stall warning horn sounding. With a significant loss of engine power, he attempted to maintain 70 knots but was unable, and he made small pitch adjustments to stay above 65 knots, eventually retracting the flaps. Unable to maintain altitude, he maneuvered for an off-airport forced landing during which the airplane impacted a tree and then the ground. A postcrash fire erupted. The pilot exited the burning airplane but returned to rescue the right front and left rear seat passengers (both minors) who were unable to release their restraints for undetermined reasons. Although the pilot reported the fuel selector was on the right fuel tank, it was found selected to the left fuel tank; however, this likely did not contribute to the partial loss of engine power as both tanks were fueled before the flight. Examination of the engine, engine systems, and the remains of the left and right fuel supply and vent systems revealed no evidence of any preimpact mechanical malfunctions or failures that would have precluded normal operation. The atmospheric conditions at the time of the accident were conducive to the development of serious carburetor icing at glide power. Given the evidence, it is likely that following the prolonged wait with the engine at a low power setting before takeoff, the engine developed carburetor ice during the subsequent takeoff, which resulted in the partial loss of engine power during takeoff. Although the pilot reported a normal rotation point, the witness-reported rotation point and onboard recorded data showed the airplane’s takeoff roll was between 34% to 41% longer than the calculated takeoff roll distance for the environmental conditions that day. The longer takeoff roll and the abnormal engine noises reported by the witnesses should have alerted the pilot to the partial loss of engine power and prompted him to abort the takeoff, which would have avoided the accident.

Multiple pilots who had flown the airplane in the days leading up to the accident reported no issues with the airframe or engine. One pilot who flew the airplane 4 days before the accident reported magneto drops of about 75 rpm from each magneto during the engine check before takeoff and a maximum rpm of 2,700 during the takeoff roll. Another pilot who flew the airplane 3 days before the accident reported magneto drops within 25 rpm of each other, which was “satisfactory” and within the limits specified by the Pilot’s Operating Handbook. The airplane was not equipped with an engine monitor or a carburetor temperature gauge. A portable ADS-B transceiver and a tablet computer were recovered from the wreckage. The portable ADS-B transceiver did not have data recording functionality. The tablet was accessed, and the ForeFlight application was running in the background. Validated recorded data for the accident flight from ForeFlight totaled about 1 minute 26 seconds, beginning at 1635:11 when the flight was near the displaced threshold for runway 28L, and ending at 1636:37, when the flight was near the final resting position. The downloaded data indicated that takeoff was initiated using the displaced threshold, and the airplane was about 1,520 ft from the runway 28L threshold when it attained 59 knots groundspeed or 64 knots indicated airspeed based on the headwind component. That location was on the runway 28L centerline and right of the runway 1R centerline. According to the airplane’s maintenance records, the left and right front seat restraints were replaced with Anjou Aeronautique 3491423-12-070 16 G restraints at the airplane’s last annual inspection to comply with Airworthiness Directive (AD) 2003-26-06. The rear seat restraints, which were Pacific Scientific part numbers 0108168-11 and 0107119-55, were not changed as they were not affected by the AD. According to the Pilot’s Operating Handbook, at design maximum gross weight, the environmental conditions that existed at the time of the accident, and the pilot-reported flight configuration of takeoff flaps, the takeoff roll and distance to clear a 50-foot obstacle were about 1,135 ft and 1,765 ft, respectively. Based on the pilot-reported weight and balance calculations, the airplane was about 12 pounds under design gross weight at the start of the flight. A note in the performance section of the Pilot’s Operating Handbook specified that the distances were to be reduced by 10% for each 10 knots of headwind. After an uneventful flight on December 14, 2020, the airplane was fueled per the club policy, and 8.51 gallons of 100 low lead fuel were added. The airplane had not been operated between the conclusion of the flight on December 14, 2020, and the accident flight. According to the manager of the fuel facility that supplied the fuel for the airplane, there was no water contamination of their fuel found either by visual inspection of a sample or from a water sensor installed in the tank. Additionally, there were no reports of any fuel related issues from any of the aircraft fueled from the same source. Examination of the accident site revealed the outer section of the left wing with attached aileron was located in the first or primary impacted tree; the tree trunk was fractured about 8 ft above ground level. The main wreckage consisting of the fuselage with attached right wing and empennage came to rest upright near the base of a large tree adjacent to a road. The fuselage came to rest heading about 180° opposite the direction of flight about 3,900 ft west-southwest from the departure end of the runway 28L. The cockpit and cabin portions of the fuselage were nearly consumed by a postcrash fire. Examination of the flight controls for roll, pitch, and yaw revealed no evidence of preimpact failure or malfunction. The flaps were in the retracted position based on the position of the flap motor. Fuel consistent with 100 low lead was noted in the right fuel tank with no evidence of water or contaminants; the amount was not quantified. The left fuel tank was ruptured with no fuel remaining. Examination of the right fuel vent system revealed it was free of obstructions from the end of the vent tube at the bottom of the wing into the fuel tank. The right fuel supply system was continuous from the tank to the fuel selector valve to the auxiliary fuel pump; the line between the auxiliary fuel pump and the engine-driven fuel pump was heat damaged. Examination of the left fuel vent system revealed the vent line was burned and exhibited internal contamination that was consistent with organic material. The left fuel supply line was damaged between the fuel tank and the selector valve. Examination of the cockpit and cabin revealed the fuel selector was positioned to the left fuel tank. The fuel selector sustained heat damage but was internally free of obstructions. Small particles consistent with metallic shavings were noted in the valve filer bowl. Both front seats and the rear bench seat remained attached to the structure. A buckle assembly found in the right rear portion of the cabin and a buckle assembly with attached connector link found in the left rear portion of the cabin were retained. The buckle assembly for the right front seat position was not located. Examination of the buckle assembly with attached connector link was performed by the NTSB Materials Laboratory. The buckle assembly had darkened due to fire exposure and was coated with melted aluminum from the burned aircraft structure. Melted aluminum fused the components of the buckle together, and the buckle could not be mechanically manipulated without possible destruction of the internal components. Consequently, the angle at which the link connector became disengaged from the buckle and the force required to open the buckle could not be reliably measured and compared to specifications. X-ray examination revealed the connector link appeared to align normally with the internal clips, and the amount of rotation of the link connected within the buckle also appeared normal and would not be expected to interfere with unlatching. The components of the belt buckle appeared to all be present and assembled correctly with no obvious visual damage. Consequently, there appeared to be nothing mechanically wrong with the buckle or connector link that would have precluded normal unlatching and operation. Examination of the engine, which separated from the airframe, revealed the crankshaft was fractured aft of the crankshaft flange; the separated section of crankshaft flange remained connected to the propeller flange. The engine exterior was discolored and sooty consistent with exposure to the postimpact fire. During rotation of the crankshaft using a tool inserted in the vacuum pump drive pad, crankshaft, camshaft, and valve train continuity were confirmed. Compression and suction were observed from all cylinders. Borescope inspection of the cylinders revealed no anomalies. The engine-driven fuel pump was impact fractured, and the pumping section separated from the engine. The pump’s rubber diaphragm was fire destroyed but the internal valves appeared normal. The pump’s actuator rod was observed to move normally up and down during crankshaft rotation. The carburetor remained attached to the engine and exhibited exterior discoloration and soot consistent with exposure to the postimpact fire. The carburetor’s two right side attaching nuts were secure and required normal force to remove, and the two left side nuts were finger loose. The mounting gasket exhibited gray coloration on the side where the nuts were secure and pink coloration on the side where the nuts were finger loose. No liquid drained from the carburetor when it was inverted. The throttle and mixture control cables remained attached to their respective control arms; both control cables were fractured consistent with impact. The throttle contr