When approximately 120 NM north of Williamtown, the crew of the BAe 146 aircraft received a fire warning for the number 3 engine. After confirming the indication and completing the appropriate check list items, the crew shut the engine down and the fire warning stopped.
The crew informed air traffic control of the engine shut down and of their intention to continue to Williamtown. The emergency services at Williamtown were placed on standby; the aircraft was landed without further incident.
Preliminary inspection of the number 3 engine by the operator indicated the presence of extreme temperature around the bleed-band area and a cracked fuel line between the flow divider and fuel manifold assembly. The engine had accumulated 922 hours since overhaul by its manufacturer in the USA in June 2000 and 16,470 hours since new.
The cracked fuel line was removed to be examined by the Australian Transport Safety Bureau. It carried the following identification: 91547 - ASSY - 2 - 193 - 940 - 02 and CDA 99193. The records indicated that the fuel line was fitted during the last engine overhaul.
The fuel line was a fabricated assembly with stainless steel unions that were gas-tungsten arc welded to each end of the stainless steel tubing. The fuel line had failed immediately inboard of the weld between the tube and union on the flow divider end of the unit. The examination of the fracture surfaces showed that the cracking emanated from a point on the inside bore of the tube.
The cracking was consistent with a fatigue mechanism propagating under high frequency, low magnitude vibratory loads. The fatigue crack had propagated circumferentially and covered approximately eighty percent of the cross-section. The remaining section failed in overload. The investigation found no evidence of any physical defect or prior cracking. The material of the tube and the unions complied with the manufacturer's material specifications.
The fuel line incorporated an expansion loop to reduce the assembly and operating stress. The issue of these stresses contributing to fuel leakage at the unions has been previously identified and was addressed by Textron Lycoming service bulletin ALF 502R 73-2 Revision 1 of March 1992. The service bulletin also stressed a need for the fuel line to be positioned so as to preclude stresses before tightening the unions and securing clamps.
While these issues may have contributed to the cracking, the investigation was unable to conclusively determine the reason for the fatigue cracking of the tube.