WPR24LA134 · Bell 206L-3

During a pipeline inspection flight, the pilot and passenger of the accident helicopter heard a loud noise, followed by a shudder and an uncontrolled descent. The pilot was unable to stop the helicopter’s descent with collective control inputs and he completed a forced landing into an orchard, which resulted in substantial damage to the tailboom. Postaccident examination of the helicopter revealed fractures in the upper and lower arms of the cylinder extension to the collective control servo. All the linkages and securing hardware for the collective control system were in place. Metallurgical examination revealed the collective servo arm fractured due to fatigue cracking that had initiated in the lower cylinder extension arm at corrosion pits along the inside surface. Once the crack had propagated through more than 80% of the arm’s cross section, the remainder carrying the loads fractured from overstress. The upper arm also fractured from fatigue initiating at corrosion pits along the radii at the forward and aft ends of the open section in the middle of the arm. Composition testing determined that the elements in the corrosion pitting were consistent with compounds found in marine environments. The helicopter had been periodically flown about 50 nm from salt water. However, the undamaged cyclic servo had also been installed in the accident helicopter at the same time as the collective servo. It is unclear why the cylinder extension arms of the collective servo failed, and the cyclic servo cylinder extension arms remained intact, both having been exposed to the same environmental conditions; the extension arms for both the collective and cyclic servos were contained within a protective housing above the cockpit. The fracture surfaces were also examined to count the fatigue striations and correlate the fatigue crack growth with the 59 hours that the helicopter operated with newly installed composite blades. However, the number and pattern of crack arrest features could not be linked to any specific loading events, and the linking of the individual striations to the two-vibrations-per-blade revolution was inconsistent with the 59-hour propagation time. Therefore, the fatigue crack initiation or propagation could not be attributed to the blade installation.

Verbatim NTSB analysis and probable cause from the NTSB dataset