Accident Details
A twin-engine aircraft, registration C-FCAS, was conducting a calibration flight on behalf of Transport Canada. The flight departed Montreal-Dorval Airport at 1741 local time to calibrate a radar antenna. After establishing contact with air traffic control, the crew began an orbit 20 nautical miles from the antenna at an altitude of 1,700 feet. Following the crossing of the 165 radial, the aircraft entered an uncontrolled descent and crashed in a field near Sherrington. The aircraft was totally destroyed, and both occupants were killed.
Investigation Findings
Examination of the wreckage revealed that the outboard section of the right wing, beyond the right engine nacelle, had separated in flight. This separated section was found approximately 640 meters from the main wreckage. Laboratory analysis determined that the right-hand lower forward outboard wing attachment fitting of C-FCAS had developed a progressive crack. This crack was of sufficient magnitude to cause in-flight fracture under normal flight loads, leading to immediate wing separation.
The analysis indicated that the progressive cracking was complex and involved mixed modes in the initiation and early growth stages. Later stages were predominantly fatigue cracking driven by cyclic stresses generated in flight. The failed fitting exhibited manufacturing deficiencies in the form of gas bubbles located along grain boundaries, which strongly influenced the early cracking. It was not possible to precisely evaluate the mode or rate of crack propagation due to insufficient data on the adverse influence of these gas bubbles on material performance.
The fitting met all relevant specifications. Investigators concluded that typical quality control procedures would not have detected the gas bubble deficiency. The available evidence suggested that only some fittings from a relatively small production batch of part number 50-110057 were defective in this manner, and similar cracking occurred only in that batch.
Inspection History
The fitting was inspected in accordance with Airworthiness Directive 70-25-04 and Service Instruction 0394-018 at 4,907 hours, and no crack was detected at that time. Following the accident, the FAA and the manufacturer affirmed that a dye penetrant inspection should have been performed at 5,407 hours, based on a 500-hour inspection interval. This inspection was not conducted. Had it been performed, it was most probable that a crack would have been detected. A lack of clarity in AD 70-25-04 led maintenance personnel, including those from the Department of Transport, to interpret the directive as making the 500-hour dye penetrant inspection non-mandatory if wing skin cracks were not found during regular 100-hour skin inspections, which had been carried out.
Probable Cause
The following findings were identified: The right-hand lower forward outboard wing attachment fitting of C-FCAS had developed a progressive crack of sufficient magnitude to cause in-flight fracture and consequent immediate wing separation outboard of the right engine nacelle under normal flight loads. The laboratory analysis revealed the progressive cracking was complex and of mixed modes in the initiation and early growth stages, with later stages essentially due to fatigue cracking driven by cyclic stresses. The failed fitting exhibited a manufacturing deficiency in the form of gas bubbles located on the grain boundaries, which strongly influenced early cracking. It was not possible to precisely evaluate the mode or rate of crack propagation due to insufficient data on the adverse influence of gas bubbles. The fitting met all relevant specifications, and typical quality control would not have detected the deficiency. The evidence indicated that only some fittings in a small production batch were defective, and that cracking similar to that in C-FCAS only occurred in this batch. The fitting was inspected per AD 70-25-04 at 4,907 hours with no crack detected. Since the accident, the FAA and manufacturer affirmed that a dye penetrant inspection should have been made at 5,407 hours based on a 500-hour interval; this inspection was not made, and had it been made, a crack would most probably have been detected. There was a lack of clarity in AD 70-25-04 that led maintenance personnel to interpret it as making the 500-hour dye penetrant inspection non-mandatory if wing skin cracks were not found at regular 100-hour inspections.