ERA12LA493

The pilot's failure to recognize that the helicopter was experiencing tail rotor dynamic instability and to take immediate corrective actions during an external load lift, which resulted in the failure of a tail rotor blade.

The pilot had flown two uneventful external load lifts to place 2,900-lb air conditioning units on a warehouse roof. The pilot reported that, during the third lift, he felt vibration in the pedals that became violent, and the helicopter then began to rotate about its vertical axis. The air conditioning unit touched down on the roof as the helicopter was spinning. The pilot could not stop the helicopter's rotation, so he released the cable attached to the air conditioning unit and then maneuvered the helicopter away from the warehouse. He then increased forward speed, turned right to line up with a street, and conducted a roll-on landing. Examination of the helicopter revealed that the entire aft portion of one of the four tail rotor blades had separated just aft of the blade's spar where a bond line existed. Examination of the tail rotor blade revealed high-stress progressive crack growth features at the root end of the fracture, buckling deformation adjacent to the fracture, and bending deformation of the leading edge, all of which were consistent with the tail rotor blade fracturing due to dynamic instability in the tail rotor. The progressive crack growth features observed on the fracture surface were associated with relatively high stress and few cycles and likely occurred after the deformation associated with the buckling. In addition, all four tail rotor blades exhibited bending deformation, indicating that they all experienced loads that exceeded the allowable design loads, and the deformation pattern was consistent with an external input on the tail rotor assembly overloading all of the blades rather than a failure in the blade causing it to become unstable. The helicopter manufacturer confirmed that such damage can be caused by dynamic tail rotor instability and that such instability can be accompanied by tail rotor vibration, as was experienced during the accident flight. Although dynamic tail rotor instability rarely occurs, it has been known to occur on the accident helicopter make and model. To improve tail rotor stability, the helicopter manufacturer had introduced two modifications to the tail rotor system, and both of these modifications had been installed on the accident helicopter. Even with the modifications, dynamic tail instability can occur, and high values of left pedal, improper tail rotor cable tension (too high or too low), bottoming of the tail rotor control system spring, higher rotor speed, or relative wind from the right forward quadrant could increase susceptibility. However, a review of the helicopter's maintenance records did not reveal that any of the mechanical factors that could contribute to tail rotor instability existed, and wind was calm at the time of the accident. Additionally, after the damaged components were replaced, the helicopter was returned to service. The helicopter's flight manual also contained guidance stressing that pilots should immediately decrease the tail rotor pitch after encountering pedal vibration. If the pilot had recognized that the pedal vibration was indicative of tail rotor instability and immediately taken the proper corrective actions in accordance with this guidance, the accident might have been prevented.

Aircrane Inc. was established in 1993 as a construction helicopter operator, specializing in Heavy lift Aerial Crane Services or "External Loads." At the time of the accident they held 14 CFR Part 133 (External Load), 14 CFR Part 135 (Air Taxi), and 14 CFR Part 137 (Agricultural Application) certificates. Tail Rotor Control System The tail rotor was controlled through a hydraulically boosted cable system with push-pull rods which connected a rear fuselage bell crank to the rotor. The hydraulic servo operated on the tension difference between the two cables and, as the system was a boosting system with the power piston in series with one of the cables and not a fully powered system, the yaw pedals had a direct mechanical link to the rotor blade pitch change mechanism. The cables were rigged to a specific tension and a spring inserted in one of the cables had as its main function the tuning of the rate of the whole tail rotor system to avoid unwanted resonances. The spring also maintained the tension over a wide range of ambient temperatures. Tail Rotor Blades The tail rotor blades were of aluminum alloy construction. The structural supporting member of the blade assembly consisted of a solid spar around which the skin was wrapped and bonded. The skin was bonded together at the trailing edge and formed an integral part of the blade structure. An aluminum foil honeycomb core was sandwiched and bonded between the top and bottom skins and the trailing edge side of the spar to form structural support for the skin. The tip of the blade was sealed by means of a riveted tip cap, and the root was sealed with cemented balsa filler. The root end of the blade assembly was also reinforced by a strap which was wrapped and bonded to both sides and around the leading edge of the blade. Review of the helicopter's maintenance manual revealed that the tail rotor blades had an unlimited life, provided that the following flight restrictions were complied with: 25 knots maximum sideward flight, minimum 10 seconds hovering turns (360 degrees), and minimum 88 percent Nr (main rotor rpm) on all taxi turns. Review of the maintenance records did not reveal however, whether the helicopter ever exceeded any of the specified flight restrictions nor could it be ascertained if a robust mechanism had ever been set up by the manufacturer or operators of the S58 that would capture these types of exceedances. Examination of the Tail Rotor Blades As part of the examination, the tail rotor blades were lettered from A to D in sequence in the direction of tail rotor rotation such that each blade trailed the next higher blade, and blade D trailed blade A. Blade A was fractured with most of the airfoil separated from the spar. Blades B, C, and D were intact. All of the tail rotor blades were Sikorsky part number 1615-30100-045. According to component log cards, all blades were installed on June 28, 2011 at 162.3 hours prior to the accident. The component log cards stated that prior to installation on the accident helicopter, blade B was last removed from another helicopter in 1993 for painting, and blades C and D were last removed from another helicopter in 1990 for vibration troubleshooting. The prior installation history for blade A was not noted on the component log. At the time of failure, the component log stated blade A had a total time of 2,494.40 hours. The total times for blades B through D were unknown. Data plates affixed to the inboard sides of the blades indicated the blades had been inspected and repaired at Sikorsky. Blades C and D were each marked inspected and repaired in May, 1979. Blades A and B were marked inspected and repaired in September, 1980, and in May, 1983, respectively. The data plate for blades A and C listed total times of 2,562.10 hours and 0.0 hours, respectively. The hours for blades B and D were marked unknown. A stainless steel wear strip covered the leading edge along nearly the entire length of the airfoil back to approximately 1.44 inch from the leading edge. The wear strip is bonded to the skin with Scotch-Weld AF 30 structural adhesive film manufactured by 3M. The intact areas of the blades were initially examined for paint condition, dents, and other anomalies. As shown in figures 1 and 2, the paint was eroded away from the leading edges of the blades. The paint erosion on blade A was less than that of the other blades. A dent was observed on the outboard side of blade B at a location approximately 14.5 inches from the butt end of the blade and is indicated in figure 1. This dent, measuring approximately 1.75 inches in diameter, was the largest and deepest dent observed on the 4 blades. Smaller and shallower dents were observed on other areas of blade B and on blades A and D. A slight bulge was observed on the inboard side of blade B near the trailing edge of the leading edge wear strip. The bulge was approximately 1 inch long and was located approximately 36.25 inches from the hub end of the blade. Blade A was fractured into 2 pieces. One piece included the intact main spar, and the other piece which was recovered from the roof contained most of the airfoil. The spar was bent with the tip displaced toward the leading direction and outboard relative to the hub end. The fracture in blade A intersected the hub end of the airfoil at a location approximately midway between the spar and the trailing edge. Along most of the length of the blade, the skin was fractured at the spar trailing edge on both the inboard and outboard sides of the blade. At the blade tip, a flange bonded to the trailing side of the spar was fractured and showed flat fracture features. The tip cap on blade A was removed to facilitate examination of the fracture surfaces near the tip of the blade. When the tip cap was removed, it was noted that no lock wire was installed on the bolt attaching the tip weights. The fracture features of the flange at the blade tip were generally flat, and edges at the hub end of the flange piece attached to the spar were bent outward toward the tip. The leading edge wear strip was intact and remained attached to the piece of the skin that wrapped around the leading edge spar. The trailing edges of the leading edge wear strip had a wave pattern deformation. The wear strip was disbonded from the pieces of the skin on the inboard and outboard sides of the separated trailing airfoil piece of the blade. The fracture was mostly an adhesive fracture at the interface between the skin on the trailing piece and the adhesive that remained bonded to the wear strip. Similar features were observed along the entire length of the blade where the skin was disbonded from the wear strip on both the inboard and outboard sides of the blade. The adhesive was teal green in color and was impregnated with an open-weave fiber mesh. Portions of the adhesive appeared to be stained brown. Data sheets for Scotch-Weld AF 30, the leading edge wear strip adhesive specified in the engineering drawings for the blade assembly, state that Scotch-Weld AF 30 is an unsupported structural adhesive film. According to a technical representative for 3M, their unsupported adhesive films such as Scotch-Weld AF 30 do not have a fiber mesh. The skin fractures were examined visually. The skin fractures were all on slant planes, consistent with ductile overstress fracture and closer examination of the blade surface in close proximity to one of the fractures near the root end of the airfoil revealed that the outboard skin was bent consistent with compression buckling. The surface adjacent to the fracture on the inboard side of the blade was relatively straight which was consistent with the tension side of a bending fracture. Sections of the fracture surfaces at the blade tip and the root end of the airfoil were cut from the rest of the blade and examined using scanning electron microscopy (SEM). The flat fracture features of the fl