Flight Safety Australia September-October 2004

reduction in power from the engine. Engine Condition Trend Monitoring (ECTM) Program data from the left engine indicated that a potentially safety-critical problem existed in that engine for several weeks prior to the accident. For a variety of reasons that evidence was not detected. The pattern of evidence suggests that temperature-related damage to the left engine’s compressor turbine blades, probably due to a problem with the efficiency of the cold section of the engine, resulted in the failure of one of the blades. As part of maintenance arrangements prior to the accident, the operator had been sending the ECTM data to the engine manufacturer’s field representative for analysis, but it was not being recorded or submitted for analysis as frequently as required by the engine manufacturer, or CASA’s Airworthiness Directive AD/ENG/5 and there were deficiencies in the operator’s maintenance scheduling processes. AD/ENG/5 enabled time between overhaul extensions under less restrictive circumstances compared with those required by the manufacturer, but CASA’s surveillance system was not sufficiently rigorous to ensure that the mitigators introduced were effective. The investigation also found that the CASA system for approving maintenance organi- sations and maintenance controllers did not appropriately consider the maintenance organisation’s resource requirements. The engine failure occurred during a critical phase of flight, just prior to, or at about, the time of take–off. Takeoff speed when the aircraft became airborne was probably close to minimum control speed (Vmca) of 90 kts, not sufficient to allow the aircraft to accelerate to the best one-engine inoperative rate of climb speed (Vyse) of 107 kts with an engine failure. With an engine failure or malfunction near Vmca, the safest course of action would be to reject the takeoff. This means that the aircraft may overrun the runway and perhaps sustain substantial damage, but the consequences associated with such an accident will generally be less serious than a loss of control after becoming airborne. The operator’s procedures did not provide appropriate guidance for pilots regarding decision speeds or decision points for use for an engine failure during takeoff. Aircraft manufacturers have provided such material, but CASA has not published formal guidance material. Further, the level of training available for emergencies in this category of aircraft during critical phases of flight and at high aircraft weights is less than desirable. The runway length, and the visual appearance of the runway and buildings beyond the runway at the time of the engine failure may also have influenced the pilot’s decision to continue with the takeoff. Toowoomba aerodrome was licensed and met the relevant CASA standards. However, runway 29 did not meet the ICAO standards in relation to the runway end safety area (RESA). ■ Details of the report and CASA’s responses can be found at http://www.atsb.gov.au/aviation/ occurs/occurs_detail.cfm?ID=315 and www.atsb.gov.au/aviation/rec/rec_detail. cfm?ID=444, 445 and 469-472 Toowoomba accident claims four lives FLIGHT SAFETY AUSTRALIA SEPTEMBER-OCTOBER 2004 55 Safety Investigator Australian Transport Safety Bureau 56 FLIGHT SAFETY AUSTRALIA SEPTEMBER-OCTOBER 2004 Australian Transport Safety Bureau Collision on final approach Occurrence 200201846 At about 1525:34 Eastern Standard Time (EST) on 5 May 2002, a Piper PA-28-161 aircraft, registered VH-IBK, and a Socata TB-9 aircraft, registered VH-JTV, collided on final approach to Bankstown airport. General Aviation Airport Procedures (GAAP) were in operation at the time, and aircraft were using simultaneous contra- rotating circuits onto parallel runways in the 29 direction. Under GAAP, pilots were responsible for separation when airborne in the circuit, and air traffic controllers were responsible for runway separation and the provision of sequencing and traffic information to pilots. The primary mitigator in place at Bankstown to prevent this type of collision was the provision by controllers of traffic information to pilots of aircraft operating to parallel runways spaced less than 210 m apart. The investigation concluded that this procedure reduced collision risk, but due to limitations with its general nature and its implementation, the extent of the mitigator’s influence on collision risk could not be determined. The investigation also concluded that there were insufficient visual cues available for a pilot in one circuit to reliably assess the collision potential of an aircraft in the opposing circuit. In such situations there were also insufficient cues available for the controllers to detect a potential collision in time to provide one or both pilots with information to initiate effective avoidance action. ■ Collision with power line Occurrence 200401181 On 31 March 2004, at 0600 Eastern Standard Time, the Robinson R22 helicopter with one pilot on board departed on a ferry flight to commence mustering operation from a property approximately 15 minutes flying time to the north-north-east. The pilot reported that he landed at the property and picked up a passenger who was going to show him the paddocks, fences and laneways. At about 1000, after mustering cattle into a small paddock, the pilot made an approach to land beside a fence. The pilot indicated that the weather was fine with good visibility. The wind was from an easterly direction at 5 kts and the temperature was 25 degrees C. He reported that he saw a powerline and aimed to land adjacent to a point where the powerline changed direction. However, he did not see a third wire, which the helicopter struck at a height of about 30 ft. The helicopter spun into the ground and landed on a barbed wire fence. The helicopter sustained substantial damage to the tail boom, lower vertical fin and tail rotor blades. The two occupants escaped with minor injuries. ■ Loss of separation standards Occurrence 200205540 A de Havilland DHC-8-315 (Dash 8) was southbound to Mackay from Townsville and a Boeing 717-200 (717) was northbound to Mackay from Brisbane. Both aircraft were scheduled services and were conducting regular public transport operations. They were being radar monitored by Brisbane Centre air traffic control and the crews had been cleared to descend their IFR aircraft to 6,000 ft. The Dash 8 was in cloud and was cleared by the aerodrome controller (ADC) at 1506:17 to descend to 4,700 ft. Separation between the Dash 8 and 717 was no longer assured and there was a technical breakdown of separation. The breakdown occurred because the aircraft were no longer under radar control and no procedural separation standard had been established. The breakdown of separation was not recognised by the ADC. At 1507:53, the crew of the 717 reported visual at 7 NM and advised that they were able to track for left base. The ADC told the crew to maintain 3,000 ft and to track for left downwind. The crew of the Dash 8 reported visual and were subsequently cleared to make a visual approach straight in to runway 14 at 1508:38. At 1508:51, the crew of the 717 was cleared for a visual approach. The 717 crew responded by reading back the clearance. The ADC then said 'and maintain downwind heading, the Dash 8’s currently at 9 miles'. The 717 crew neither heard nor responded to this subsequent transmission by reading back of the requirement. Approximately 90 seconds later, the ADC observed the 717 turning left base into conflict with the Dash 8 on final approach. This was a second, and more critical, breakdown of separation between the two aircraft. ■ Safety briefs • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • FLIGHT SAFETY AUSTRALIA SEPTEMBER-OCTOBER 2004 57 Australian Transport Safety Bureau Compressor stall after takeoff Occurrence 200300040 The crew of a SAAB 340B, VH-EKN, on a scheduled flight from Orange to Sydney, NSW, reported that shortly after takeoff, as they were setting climb power, they heard a ‘bang’ similar to the sound of a compressor stall. The left engine gauges indicated zero torque and excessive inter- turbine temperature (ITT). The left over temperature and ignition lights illuminated followed by a master warning annunciation. The crew carried out the engine failure procedure and, having shut down the left engine, returned to Orange and landed. Eight days later, the crew of another SAAB 340B, VH-OLM, operated by the same company, reported that shortly after takeoff from Orange, the right engine displayed characteristics consistent with a compressor stall (ATSB Occurrence 200300078). On that occasion, after the crew carried out the appropriate checklist procedures, normal engine operation and indications were restored and the flight continued to Sydney without further incident. The affected engines from both aircraft were removed for examination at the manufacturer’s overhaul facility. The engine manufacturer, after reviewing the recorded engine data, identified a number of similar conditions had existed during both flights that may have affected the engines. Both events had occurred on the first flight of the day. On both occasions a significant temperature inversion existed at approximately 1,000 ft above ground level, and the compressor stalls occurred when the crews were adjusting the power setting to climb power. The engine manufacturer’s assessment concluded that a combination of environmental conditions and engine operating procedures had led to both events and made several recommendations to the operator. Those recommendations were actioned by the operator and there have been no further instances of compressor stalls in the climb after takeoff. ■ Loss of torque Occurrence 200205705 An abnormal noise coming from the left engine of an Aerospatiale AS.332L Super Puma helicopter, registered VH-BHY, was reported by several operating crews over the course of successive charter flights. This noise and vibration was evident at approxi- mately 26,000 rpm during acceleration and deceleration phases, and could be induced by rapid movement of the speed select lever. Although the engine was operating within the manufacturer’s vibration limit, maintenance personnel continued to investigate the cause of the vibration. It was discovered that the six attachment bolts at the engine end of the Bendix driveshaft had worn through approximately 50 per cent of their thickness. There was evidence of fretting of the mating surfaces of the Bendix and engine coupling plates. The engine attachment plate for the Bendix also incurred severe elongation of the bolt holes. The Bendix shaft had completed approximately 380 hours in service since installation. New Bendix attachment bolts, along with several of the damaged bolts were forwarded to the ATSB for laboratory analysis. The examination found no evidence of material deficiencies or other anomalous features of the bolt construction that would have rendered the items susceptible to the type of damage observed. The ATSB concluded that the damage sustained by the Bendix attachment bolts supplied was consistent with the effects of inadequate bolt tightness. The engine was shipped to the manufacturer in France for further testing. The damage to the engine, observed by the manufacturer and the BEA, in their opinion was consistent with the findings of the ATSB. ■ One-engine inoperative landing Occurrence 200203655 The Saab 340B aircraft was on descent for a landing at Sydney Airport. During the descent the aircraft suddenly yawed to the left and the left propeller feathered. The crew reported illumination of master warning and caution lights consistent with a left engine failure. The aircraft’s auto-coarsen system was selected to ON and the ignition switches were in NORM throughout the incident. The auto-ignition system did not operate. The crew declared a PAN to air traffic control and an uneventful one- engine inoperative landing was conducted. The operator examined the aircraft and replaced the left engine’s hydromechanical unit (HMU) and the digital electronic control unit (DECU). Those components were returned to the repair vendor for examination. The component repair vendor found no problems with the HMU and DECU and both components were returned to the operator. The DECU was then fitted to another aircraft in the operator’s fleet and was subsequently removed following engine torque fluctuation problems on that aircraft. A subsequent examination of the DECU by the repair vendor following that incident, revealed an internal fault. The engine manufacturer indicated that DECU torque fluctuation problems had caused uncommanded auto-coarsen events in other aircraft. No flight data recorder information was available for the original incident due to a data recorder problem. The lack of flight data recorder information did not allow a detailed examination of all of the issues surrounding the incident. The reason for the apparent engine failure was not able to be fully determined. The ATSB found that some of the circum- stances were consistent with an inadvertent auto-coarsen event. Two local safety actions were identified during the investigation. One relating to the engine manufacturer’s testing procedures for the DECU in the event of torque fluctuation problems, and another relating to the operator’s Operations Manual procedures. ■