Sikorsky S61 N908CH

During the approach to the airfield at Camp Dwyer, the helicopter entered an uncommanded left yaw while approaching its destination. During the subsequent emergency landing, the helicopter impacted the ground and rolled on its right side, resulting in serious injuries to the three occupants and substantial damage to the helicopter. Investigation found there was no evidence of a preimpact structural failure, nor a main or tail rotor system failure, nor a malfunction of either engine. The image recorder installed in the helicopter showed that, about 9 seconds before the end of the recorded data, the left seat pilot’s left pedal suddenly moved to the fully forward position without pilot input, which caused the helicopter’s left yaw. The left pedal remained in its fully forward position, and the helicopter continued to yaw to the left for the remainder of the recording. Examination of the auxiliary servo cylinder assembly (part of the helicopter’s directional flight control system) found fatigue cracks on the housing of the yaw channel pedal damper check valve and the housing bolts. These fatigue cracks initiated before the accident flight and propagated until one of the bolts failed in overload, which unseated the check valve housing, allowed pressurized hydraulic fluid to escape from the upper side of the pedal damper piston (as evidence by the extruded O-ring at the check valve housing and the small pool of hydraulic fluid on the airframe structure next to the auxiliary servo cylinder yaw channel), and caused the piston to move upward. This upward movement resulted in the uncommanded full left pedal movement in the cockpit and a resultant increase in tail rotor thrust, causing the helicopter to yaw left. Although the investigation was unable to determine if the flight crew attempted to press the right pedal after the onset of the left yaw, crew movement of the right pedal would likely have been difficult due to the presence of residual hydraulic pressure within the pedal damper piston, which led to the uncommanded full left pedal movement. Thus, the pilots had limited available options to slow the left yaw. After the onset of the left yaw, the right seat pilot set the speed selector levers (engine throttles) to idle, which reduced engine power to the rotor system, reduced main rotor torque, and substantially increased the left yaw rate (as observed in the image recorder data). The emergency procedures for a tail rotor malfunction called for the speed selector levers to be set to idle assuming that the malfunction was causing a right yaw, which would be experienced during typical tail rotor malfunctions, such as a loss of tail rotor drive. The accident pilots recalled that they heard a bang and felt a shudder. The helicopter initially yawed left with a slight roll to the right, and soon after the left seated pilot stated, “let’s get this down on the ground”. The pilots stated they started emergency autorotation procedures and had no pedal or cyclic authority, and recalled the helicopter subsequently spinning to the right. However, the cockpit image recorder and data showed no change in the direction of the helicopter’s yaw to the left; however recorded data showed an increase in right roll as the helicopter descended. The pilots’ action to initiate autorotation led them to reduce engine power, but this action exacerbated the left yaw, which continued until ground impact. After the initial loss of yaw control, the helicopter also experienced large excursions in the pitch and roll axes. The excursions in pitch and roll, as evident in the recorded angular data and acceleration data, could have affected the pilots identification of the yaw direction after the emergency autorotation procedures were initiated. The helicopter’s high left yaw rate, high nose up pitch attitude, and right roll angle resulted in an uncontrolled ground impact. The pedal damper check valve conformed to drawing requirements except that the edge where the fatigue crack initiated, which had a radius of about 0.003 inches, did not conform to the drawing requirement for all sharp edges to have a radius between 0.005 and 0.015 inches. The nonconforming edge break was likely a factor in the initiation of the fatigue crack on the pedal damper check valve housing bolt lug. However, given the large area of stable fatigue crack growth on the pedal damper check valve housing, the loads on the pedal damper check valve housing bolt lug were likely low. Thus, the nonconforming edge break was likely not the only factor that led to the initiation of the fatigue crack. Before this accident, Sikorsky was aware of five events involving cracks and fractures of the pedal damper check valve housing and its bolts. Each of these events resulted in an uncommanded yaw from which the flight crews were able to recover. Sikorsky’s investigation of these events determined that improper torque of the pedal damper check valve housing bolts was the primary factor that led to these events. As a result, this investigation considered whether the pedal damper check valve housing bolts were improperly torqued during the last overhaul of the auxiliary servo cylinder, which occurred about 2.5 years and 1,270 hours before the accident. However, examination of the bolts found no evidence indicating that they had been over or undertorqued. The investigation could not determine, from the available records, when the accident check valve housing was installed onto the auxiliary servo cylinder assembly. As a result, the total accumulated time of the pedal damper check valve housing was not known. The pedal damper check valve housing had no life limit and was not replaced during the last overhaul of the auxiliary servo cylinder. During the last overhaul, a fluorescent penetrant inspection was performed to detect fatigue cracks initiating at the surface of the housing. No cracks or fractures were found; thus, the fatigue crack on the pedal damper check valve housing and its bolts initiated after the last overhaul of the auxiliary servo cylinder assembly. Nevertheless, the addition of a life limit for the pedal damper check valve housing could reduce the possibility of fatigue crack initiation during operation. At the time of the accident, there was no specific inspection for the pedal damper check valve housing and its bolts after the installation of the auxiliary servo cylinder onto the helicopter. The required safety inspection, occurring every 15 hours, and the required phase V inspection, occurring every 150 hours, both comprised a general visual inspection of the rotor flight controls. The pedal damper check valve would not be readily visible during these generalized visual inspections due to the installed position of the auxiliary servo cylinder assembly. Further, the check valve housing bolts would likely appear to be installed properly unless the auxiliary servo cylinder was removed from the helicopter and the bolts were checked using a torque wrench. Thus, the inspection guidance at the time of the accident would not likely readily identify fatigue cracks on the pedal damper check valve housing and its bolts. Inspections specifically tailored to the pedal damper check valve housing would most likely increase the probability of finding fatigue cracks. On October 17, 2022, Sikorsky released an alert service bulletin that addressed the inspection of the auxiliary servo cylinder pedal damper check valve housing. In addition, Sikorsky implemented a daily inspection of the check valve housing and a 30,000-hour life limit for the check valve housing

Quoted verbatim from the NTSB record.