WPR18FA131 · American Air Racing Ltd THUNDER MUSTANG

The airline transport pilot of the high-performance air racing airplane was taking part in an in-flight photography mission with two other airplanes. About 1 hour into the second flight of the day, the group was getting fatigued and decided to return to the departure airport. As the airplanes approached the airport, the accident pilot transmitted a "mayday" call and reported that he was going to land on the runway located directly ahead of his position. The airplane began to descend while performing an S-turn, and touched down beyond the midpoint of the 9,000-ft-long runway at a slightly higher speed than normal. A 4-knot tailwind prevailed on the landing runway about the time of the accident.

A 1,200-ft-long series of propeller strikes on the runway were consistent with the pilot applying heavy braking after touchdown. The airplane veered right as it reached the end of the runway, entered a gravel area, nosed over, and came to rest inverted. Given the airplane's nominal landing distance in addition to the factors that increased that distance on the accident landing, namely, remaining runway at the time of the touchdown, the airplane's higher landing speed, and the tailwind, the pilot would have had very little margin for error before the airplane's landing distance required exceeded the available runway.

The vertical stabilizer collapsed when the airplane nosed over, which resulted in the canopy structure contacting the ground and subsequently failing. As a result, the pilot's head was impinged at an angle against the ground, resulting in airway restriction. The pilot's extraction from the airplane by first responders took about 45 minutes, and during that time, the pilot died of asphyxiation. However, unless he had been repositioned and his breathing enabled almost immediately following the accident, survival would have been unlikely, and based on the airplane's weight and inverted position, an immediate rescue was not possible.

It could not be determined if the canopy bow was designed to be structural in nature; additionally, visual inspection revealed defects that would have further weakened its supporting properties.

Examination of the engine revealed that the coolant pump drive pulley had detached due to fatigue failure of its attachment cap screws. Separation of the pulley resulted in the detachment of the engine's two parallel serpentine drive belts, which drove multiple other engine accessories. This design allowed for a single point of failure, which resulted in a total loss of engine oil pressure, propeller governor control, and auxiliary electrical power. The belts also dislodged a coolant line, resulting in the loss of all engine coolant. With these failures, the engine would have been able to operate for a short duration before experiencing catastrophic failure, negating the pilots ability to perform a go-around, and evidence suggests that it continued to operate at a low power setting during the descent and the landing roll.

Hardness testing of the pulley attachment screws revealed that they were of the proper tensile strength. Substantial fretting damage was present on the pulley contact faces and under the screw contact areas, and thread wear was present in the pulley attachment holes. Evidence of the use of thread locking material was observed; therefore, it is likely that the screws detached due to insufficient tightening at the time of installation. Although the thread locking material used was consistent with the engine manufacturer installation instructions, product literature from the manufacturer of the thread locking material indicated that another type was available that was specifically tailored for pulley applications. Whether the use of the alternate thread locking material would have affected the outcome could not be determined.

The pilot performed all the maintenance work on the engine, which had been overhauled about 20 flight hours before the accident; however, he performed multiple significant maintenance events on the engine between the overhaul and the accident flight, so the precise timing of the pulley installation could not be determined.

The 80-year-old pilot had extensive experience in air racing and the airplane type, and had successfully dealt with multiple loss of engine power events in the airport environment and accident airplane type. However, under the specific circumstances of this failure, the design of the propeller was such that, following the loss of governor control, the propeller would have moved to a blade pitch angle that would have resulted in less drag and a longer gliding profile than the pilot had likely experienced in previous events. No logbooks of his flight experience were recovered; however, evidence suggests that he had cut back on his flying activities during the recent period leading up to the accident. The pilot was likely fatigued from the two flights on the day of the accident and appeared to be experiencing a gradual degradation in his general health, which may have begun to affect his performance.

Although there was evidence that the pilot had used two opioid pain medications at some time before the accident, active drugs and their metabolites were found only in urine. This indicates the active compounds were no longer present in his system and therefore would not have been causing any effects. Overall, there is no evidence the pilot was impaired by a specific medical condition or use of medications or other substances at the time of the accident; however, the subtle impairing effects of withdrawal from the opioid medications cannot be eliminated.

Verbatim NTSB analysis and probable cause from the NTSB dataset