ANC18LA020 · Eurocopter France SA330J PUMA J

According to the pilot, after landing the helicopter on the deck of the ship, power was increased on the No. 2 engine, while the No. 1 engine was reduced to ground idle in preparation for a freshwater rinse of the engines to prevent corrosion. Once the engines were configured, the pilot saw the lead mechanic squeeze the trigger on the wand, and then there was a rumbling sound followed by a rough shudder through the airframe. He then looked back at the cockpit instruments and everything looked normal for the condition the controls were in at the time. About that time, there was another shudder through the airframe, and he heard a change in the tone of the engine and gear box noise. He saw that the rotor speed had started to rise, and the gas generator speed of the No. 2 engine had also started to increase. He shut down the engines. Examination of the helicopter revealed that the tail rotor drive shaft had sheared, disconnecting the tail rotor from the main gear box.

Examination of the main gear box right freewheel assembly revealed that the small roller bearings were dislocated, the bearing cage was fragmented, and both tabs of the shur-lok washer showed signs of deformation with visible impact damage. Damage to these tabs is consistent with a sudden engagement of the bearing, resulting in an instantaneous spike in torque, known as freewheel jerk.

During the transition of the No. 2 engine from the idle position to the flight position, the condition of the freewheel small roller bearings likely prevented the freewheel cage from maintaining the large freewheel rollers on the ramp due to the effect of the freewheel spring. This situation is unstable until an "imbalance" modifies it and suddenly causes the cage and the freewheel rollers to rise on the ramp, imparting a torque spike into the rotor drive system.

The freewheel jerk felt during the rinsing operation of the No. 2 engine, with the No. 1 engine at idle, is likely the consequence of the degradation/dislocation of the roller bearing of the right freewheel (No. 2). Evidence of the degradation of the freewheel unit bearing was present in the two previous oil analyses. Although samples were taken before the accident in the most recent inspection, one sample was misplaced and not sent for analysis until after the event. Although not required per the maintenance manual at the time taken, had both of these samples been analyzed at the time of the inspection, it is likely the operator would have been able to detect the increase in metal before the bearing wear became severe enough to cause the jerk. After the accident, the operator implemented an internal requirement for oil sample analysis in shorter intervals.

Verbatim NTSB analysis and probable cause from the NTSB dataset