Fairchild SA226-TC, VH-EEQ

The right engine of a Fairchild Metro II aircraft, VH-EEQ, failed at the top of descent to Launceston. The aircraft subsequently completed an uneventful single-engine landing. On-site inspection revealed that the engine failure was caused by the disintegration of the second-stage turbine wheel and that segments of the wheel had penetrated the engine nacelle. One fragment penetrated the fuselage and was found lodged in the cabin lining.

The AlliedSignal TPE331-3U-304G engine, serial number P03214C, was reported to have completed 6,017.4 hours and 6,273 cycles since new.

The second-stage wheel was identified as part number 868272-1, serial number 1-01345-1313. The recorded history of the second stage wheel indicated that the wheel had been in service for a total of 2,972.84 hours and 2,878 cycles at the time of failure on 5 March 1999. The wheel was installed new in engine s/n P90105 on 11 Jan. 1982. It was removed from engine s/n P90105 on 25 Sept. 1992 with a recorded service life of 564.74 hours and 508 cycles. The wheel was installed in engine s/n P03214C on 30 Aug. 1994 with a recorded history of 564.74 hours and 508 cycles in service. There remain, however, uncertainties in the recorded service history of the part.

The uncontained failure of the second-stage turbine wheel was caused by the progressive reduction of the wheel cross-section, during operation, near the transition from the hub to web. No material anomalies or regions of pre-existing crack growth were associated with the separation of wheel segments.

Analysis of the recovered turbine components revealed that the reduction in wheel cross-section was caused by the effects of sliding contact with sections of the fractured second-stage wheel knife-edged seal. The web was reduced from original approximately 13 mm to approximately 8.5 mm.

Examination of the remnants of the second-stage seal indicated that the fracture of the seal had occurred as a result of fatigue crack growth from the radii at the corners of the slots formed in the forward edge of the seal.

A significant variation in the radii of the slot corners was noted when a comparison was made between the fractured seal and the seal from another second-stage wheel assembly. The one intact slot corner of the fractured second-stage wheel seal was sharper and did not display the smoothly rounded characteristic of the corners of other seals. A decrease in radius would create an increase in stress concentration and increase the likelihood of fatigue crack initiation.

As a result of this investigation the following interim recommendations were issued.

IR19990112

The Bureau of Air Safety Investigation recommends that the engine manufacturer, AlliedSignal, audit the process employed to manufacture TPE 331 turbine wheel knife-edged seals should be audited to determine those factors that may lead to excessive variations in slot corner radii.

IR19990113

The Bureau of Air Safety Investigation recommends that the engine manufacturer, AlliedSignal, determine the sensitivity of turbine wheel seals to the initiation of fatigue cracks from slot corners, as a function of slot corner radii.

IR19990114

The Bureau of Air Safety Investigation recommends that the engine manufacturer, AlliedSignal, consider, during engine design and the formulation of continuing airworthiness instructions, the possibility that turbine wheel seal fatigue failure may result in hazardous modes of engine failure.

IR19990115

The Bureau of Air Safety Investigation recommends that the Civil Aviation Safety Authority Australia conduct an audit of that part of the aviation safety system that establishes the operational history of life-limited components, to establish why the operational history of second-stage turbine wheel, p/n 868272-1, s/n P03214C, could not be determined with certainty.

IR19990116

The US Federal Aviation Administration should note the safety deficiency identified in this document and take appropriate action as considered necessary.