WPR14TA370 · Eurocopter AS 350 B3

The purpose of the public helicopter flight was to provide recurrent emergency procedures and night vision goggle training for the commercial pilot. Both the pilot and the flight instructor were active pilots for the law enforcement agency and current in the accident helicopter type.

Preflight checks and initial training maneuvers were uneventful. During the power recovery phase of a practice autorotation, the flight instructor applied engine power by moving the throttle twist grip from the idle to the flight position; however, the engine did not respond as expected, the rotor rpm decayed, and the helicopter landed hard. The helicopter sustained substantial damage to the tailboom; neither occupant was injured.

Immediately following the hard landing, the flight instructor observed that the rotor rpm was still low and that the amber-colored governor and twist grip warning lights, which should have extinguished during the power recovery, were still illuminated. He manipulated the throttle twist grip multiple times between the idle and flight detents in an attempt to extinguish the lights and increase the rotor speed without success. With the twist grip in the "flight" position, he then reached up and recycled the start selector switch on the roof panel, and the lights extinguished. After the flight instructor exited the helicopter and examined the damage, he got back in the helicopter, and the pilot then shut down the engine.

This was the first training flight (requiring an autorotation with power recovery) since the helicopter manufacturer had issued a service bulletin (SB), which recommended modifying the engine control logic. The SB was issued following multiple reports of engines remaining at idle power during practice autorotation power recoveries despite the twist grip being moved to the "flight" position. Although this was similar to the accident scenario, maintenance records revealed that the SB was accomplished about 46 flight hours before the accident, and postaccident examination revealed that the SB had been complied with correctly.

During the postaccident airframe examination, a small amount of play was observed in the twist grip on the flight instructor's side, which sometimes caused the governor and twist grip amber caution lights to not extinguish when the grip was in the flight detent. Wiggling the twist grip while in the flight detent resolved the issue, which the operator's chief pilot reported was not uncommon throughout the agency's fleet, and the pilots were accustomed to it. The flight instructor did not recall the status of the amber lights during the recovery phase before the hard landing. The flight procedures for autorotation training called for a confirmation that these lights were extinguished during the power recovery phase; however, the chief pilot stated that, given the minimal altitude (70 ft above ground level [agl]) that was recommended by the helicopter's manufacturer to initiate the power recovery, pilots were taught to focus on flying the helicopter and not on the lights.

Ten months after the accident, the helicopter manufacturer issued a safety information notice regarding simulated engine-off landing training, which referenced the high exposure to accidents and incidents during engine-off landings, and issued a series of procedural updates, including advising that power recoveries be initiated as the helicopter passed through 200 ft agl rather than 70 ft agl.

The status of the lights during the maneuver could not be determined; however, if they were illuminated, the engine would have been operating in "mixed" mode and would have exhibited a very slight delay in power recovery. A postaccident engine run revealed that the engine responded within specifications in mixed mode and in a series of other modes under multiple load conditions. Following the run, the engine's hydromechanical unit was removed and tested. It was slightly outside of specification for the "P3" module check, which affected operation in mixed mode. However, because the engine test run was successful, this adjustment variance was not considered causal to the engine providing insufficient power during the accident sequence. Additionally, the engine manufacturer stated that it was not unusual for this module to go out of adjustment in the field.

A failure-mode analysis was performed in an effort to determine why the engine would not resume flight power after the hard landing and whether this was related to the engine's failure to provide sufficient power when commanded by the pilot during the power recovery phase of the practice autorotation. The results revealed that the behavior could be duplicated if the throttle input lever had become declutched from the load limiter inside the hydromechanical unit. However, for this condition to have occurred, the pilots would had to have performed a highly unusual series of nonstandard procedures before the accident. It is also possible that, if declutching occurred, it could have been caused by the hard landing, in which case, it would not explain what happened in the accident. Therefore, the reason that the engine did not provide sufficient power when commanded by the pilot during the power recovery phase of the practice autorotation could not be determined.

Verbatim NTSB analysis and probable cause from the NTSB dataset