Jump to Content

Summary

Summary

After departing Brisbane en-route to Singapore, the crew of the Boeing 777-212ER aircraft heard and felt two thumps through the airframe and noticed a severe vibration indication of the right engine. The vibration subsided before re-occurring moments later with an increase in the engine's exhaust gas temperature also indicated. The crew conducted an in-flight engine shutdown and requested Air Traffic Control clearance to divert to Darwin where an uneventful single-engine landing was carried out.

An inspection conducted by ground engineers at Darwin found a stage-1, variable stator vane (VSV) control lever broken on the right engine. A boroscope inspection of the engine interior was then carried out with a number of compressor blades found damaged. The engine was removed from the aircraft and sent to an overhaul facility for disassembly and evaluation.

Engine description

The Rolls-Royce Trent 800, was a triple spool turbofan engine. Its construction consisted of a single-stage low pressure fan connected to a five-stage low pressure turbine. An eight-stage intermediate pressure compressor (IPC) connected to a single stage intermediate pressure turbine and a six-stage high pressure compressor (HPC) connected to a single-stage high pressure turbine (see fig 1).

Figure 1

To maintain maximum efficiency during all power settings, the airflow through the engine needed to be controlled to prevent stalling or surging. This was achieved by a single stage of variable inlet guide vanes installed between the fan and the first stage of the IPC and two stages of variable stator vanes, VSV stage 1 and VSV stage 2 (VSV-1 and VSV-2) installed between IPC stages, one/two and two/three respectively. Each Variable Stator Vane was attached to a lever that transferred the linear input from the controlling actuators and unison rings to a rotational movement of the vane (see fig 2). These levers consisted of an arm and connecting pin (see fig 3).

During engine start, these vanes would have been in their most closed position with internal engine bleed valves open. As the power was increased, the bleed valves would close and the vanes move toward their full open position allowing optimum airflow through the engine.

Post incident engine inspection

Prior to disassembly, the engine's exterior was inspected with the broken VSV-1 lever identified as being in the number-28 position. No evidence was observed of bird impact or other external defects. Rigging and clearance checks of the VSV control system were carried out with no anomalies found. The remaining VSV-1 and VSV-2 levers were then removed and crack tested using a dye penetrant inspection. No evidence of cracking was found on any of those levers.

After separating the engine-to-modular level, the IPC and HPC modules were completely disassembled for a detailed inspection.

IPC inspection

Removal of the IPC casing revealed six stage-2 blades displaying soft body impact damage 1 resulting in bending of the blades. On two of those blades the corners had also detached. Three other blades displayed hard body impact damage 2 with cuts and nicks (small cuts) on their surfaces. Two blades with minor nicks were found in stage 3, with only one blade in stage 5 showing nick damage. All of the stage-8 blades displayed hard body impact damage on their trailing edges, a few blades also having nicks on their leading edges. There was no evidence of damage to the disc material adjacent to the blade roots on any of the eight stages.

The IPC case lining was examined with only minor damage evident. The number-28 VSV 1 vane, had a wear mark on the leading edge lower corner with a noticeable worn stepped area on its horizontal surface above its base. The adjacent number-29 vane had a contact mark at a point mid span on the vane and one on its base. The remaining VSV-1 and all the VSV-2 vanes were found to be undamaged. Dark deposits were evident around the base of each VSV. These deposits formed a black ringed area around all except for the number-28 vane where the mark was crescent shaped.

When the number-28 and 29 vanes were positioned so that the wear marks on both vanes aligned, the number-29 vane was found to be in its normal full open position while the number-28 vane was noted as sitting beyond its normal closed position. The dark crescent area around the base of the number-28 vane also coincided with the angle of the vane's root. When the number-29 vane was moved to the closed position it was seen to nudge the number-28 vane up towards its normal closed position.

1 Having been impacted by an object made from a softer material than the blade itself.
2 Having been impacted by an object made from the same or harder material than the blade itself.

Figure 2, 3

Failed VSV-1 lever

The Australian Transport Safety Bureau (ATSB), conducted a metallurgical examination of the failed number-28, VSV-1 lever (see fig 3). The examination found that:

'The lever had fractured transversely through the end of the arm section, at a location coincident with the riveted connection to the actuator pin. The fracture path followed a uniform arc, extending from one side of the arm to the opposite and intersecting the pin connection at the centre. A prominent track mark had developed on the underside of the arm where the relative movement between the separated arm and the pin flange had produced appreciable wear. On close inspection, the fracture path appeared to intersect the bore of the rivet hole with a slight upward 'dishing' of the arm section beneath the rivet head'.

The examination also determined that:

'During riveting, the expansion of the rivet shaft could induce tensile stresses within the bore of the rivet hole if the diameter was insufficient to allow for the expansion. Tensile stresses of this nature would be expected to predispose the lever arm to the initiation and propagation of fatigue cracking'.

The examination of a further four VSV-1 levers was conducted, with welding and partial fusion between the lever and connecting pin evident, and varying degrees of cracking also evident on all four levers. For the full technical report refer to attachment A.

A further investigation by the engine manufacturer, identified the presence of a double-sided chamfer to the lever holes on a small number of levers. This removal of material during the lever manufacture may have led to the overheating and partial welding of the lever material during the rivet forming.

HPC inspection

In the HPC, all of the stage-1 blades displayed severe hard body impact damage with one blade found to have failed, detaching above the blade root. Stages 2 to 6 showed hard body impact damage to varying degrees on all the blades.

Close examination of the failed HPC stage-1 blade, found a chipped area in the leading edge, with the fracture surface revealing a number of crack progression marks indicating that the failure was progressive over a number of cycles and not instantaneous. The exact number of cycles required to fail the blade could not be determined (see fig 4).

Figure 4

Engine history

The engine commenced service in December 1998 and had completed a total of 8923 hours and 2373 cycles at the time of this incident. Its service history showed that on 8 April 2001, a routine boroscope inspection detected damage to a number of IPC stage 2 blades in the form of bending and curling to their tips. This damage was assessed to be within the manufacturer's allowable limits so the engine remained in service. On 17 October 2001, a substantial shift in the turbine gas temperature (TGT) was detected giving a warning that the engine's efficiency had deteriorated significantly. A boroscope inspection was carried out on the engine with only the previously recorded IPC stage-2 bent blade damage found. No other damage was observed on the engine. Checks of the air system, engine bleed air and monitoring systems were carried out, however the reason for the TGT shift could not be determined.

Previous VSV lever failures

Although the manufacturer had not experienced previous failures of VSV-1 levers, failures of VSV-2 levers had been recorded on two separate occasion. On those occasions the connecting pin's had fretted through the body of the lever due to inadequate riveting during the manufacturing process. The result of the levers failing was the closure of their associated variable stator vanes, which created a disruption to the airflow behind them. The vibration subsequently experienced by the blades passing the area resulted in fragments of disc material breaking off and migrating through the engine, damaging blades further down stream. These failures were indicated by a progressive increase in the engine's TGT over periods of 2 to 4 weeks.

The manufacturer issued a service bulletin, RB211-72-D516 to all operators recommending an inspection of, 'the six VSV-1 and VSV-2 levers either side of the actuating mechanism control rod connection, for significant relative movement between the lever and connecting pin'.

 
Share this page Comment