When passing through 1,000 feet after take-off from Essendon runway 26 a loud bang was heard, and the left engine was noted to have failed. Concurrent with the left engine failure the right engine driven hydraulic pump failed. The pilot initiated a return to Essendon. Due to the loss of both hydraulic systems the landing gears were extended using the manual system. The manual system does not have any provision for closing the landing gear doors therefore they remained in the down position and created considerable drag. The aircraft would not maintain height on one engine. Rather than approach and land on the shorter Essendon runways the pilot diverted to Melbourne Airport for a successful asymmetric landing.

Investigation

1. Overhaul agency investigation

The left engine was removed and transported to a manufacturer approved overhaul agency for disassembly and investigation. The engine internal rotating and stationary components were found to be severely damaged.

Assessment by the overhaul agency suggested that the primary failure was a first or second stage low pressure (LP) turbine blade. It was suggested that liberation of this blade resulted in extensive foreign object damage and imbalance of the rotating assemblies with consequent blade, shroud, disc and impellor distress.

It was determined that at the time of failure the engine had completed 6,858 hours and 4,114 cycles since new. Repairs, during which some turbine blades were replaced, were carried out in August 1989, 2,728 hours and 1604 cycles prior to this failure, and in October 1991 1,300 hours and 807 cycles prior to this failure.

2. Materials Evaluation Facility investigation

The damaged rotating and stationary components were forwarded to the Civil Aviation Safety Authority's Materials Evaluation Facility (MEF) for further investigation.

This metallurgical evaluation determined that, among other damage:

• a total of 84 1st stage low pressure turbine (LPT) blades had been fractured. The fractures were consistent with the application of excessive stress associated with interference from other turbine module components with the tips and trailing edges of the 1st stage blades while the turbine was operating.
• a total of 84 2nd stage LPT blades had fractured with the application of excessive stress. Unlike the damage to the tips and trailing edges as noted on the 1st stage blades, these 2nd stage blades were damaged on their leading edge and the blade platforms, similarly while the turbine was operating.
• the 2nd stage LPT stator assembly had been extensively damaged. The leading and trailing edges of the nozzle guide vanes had been damaged, and the shroud was fractured.

In analysing the sequence of the failure, the MEF considered:

1. the contact evidence of the aft face of the 2nd stage low pressure turbine stator shroud and the leading edges of the 2nd stage rotor blades,
2. the contact evidence of the 1st stage low pressure turbine blades with the forward faces of the 2nd stage low pressure turbine stator shroud fracture surfaces,
3. the deposits of metallic aluminium on the 2nd stage low pressure turbine stator shroud fracture surfaces,
4. the location and extent of the fracture in the 2nd stage low pressure turbine shroud, and(e) the mechanism of the fracture in the 2nd stage low pressure turbine stator shroud.

The MEF considered that the fracture of the 2nd stage LPT stator assembly was the first event in the engine failure sequence. Evidence of fatigue cracking was found at locations distributed around the circumference of the shroud. Fatigue cracking had initiated at the outer surface of the stator casing at a location that coincided with a marked change in the geometry of the shroud. Thermal expansion of the nozzle guide vanes with each engine thermal cycle would result in the imposition of alternate bending loads on the stator casing - the combination of vane expansion and restraint of the outer casing at its forward and aft edges would lead to the development of a tensile stress state at the outer surface.

3. Manufacturers investigation

The manufacturer disagreed with the MEF findings. The manufacturers analysis of the 2nd stage low pressure turbine stator shroud found no evidence of any fatigue propagation anywhere on the fracture surface. Their analysis of the secondary damage to the engine indicated it was consistent with a second stage low pressure turbine failure as evidenced by the associated damage compared to other recorded failures in the fleet. They also advise that there has not been any reported second stage low pressure stator assembly fatigue failures during more than 15 million hours of operation of this engine type.

The manufacturer considered that;

1. the heavy rub on the leading edges of the 2nd stage low pressure turbine blades was secondary to the impact damage on the leading edges of the blades,
2. the heavy rub on the aft side of the 2nd stage low pressure turbine stator had a fresh appearance and occurred after the impact damage on the vanes, and
3. the fatigue zone identified by the MEF was a shear lip with extensive smearing indicating an overload failure.

Based on their own material analysis the manufacturer stated that:

'the second stage low pressure turbine stator outer casing was not the primary failure... but rather (the primary failure was) an air foil separation of a second stage low pressure turbine blade'.

and, in relation to an air foil separation;

'Unfortunately, the exact cause of the second stage blade separations was not determined. The separation of these blades is believed to be related to the non-discernible grain areas detected in the air foils of one or more blades'.

The manufacturer used a micro etch technique to examine five random blades from this engine, finding one that had a non-discernible grain area.

The manufacturer has issued a service bulletin requiring the removal from service of a series of blades found to have non-discernible grain areas. The one blade from this engine that was identified as having a non-discernible grain area was from a series that were not listed for removal by the service bulletin.

4. Hydraulic pump failure

The failed hydraulic pump was removed and sent to an overseas overhaul agency for assessment and repair. Although it was a requirement, the overhaul agency did not supply a report on the cause of the failure.

The aircraft is fitted with two separate hydraulic systems. The fact that the right engine driven hydraulic pump failed at the same time as the left engine failed is considered to be a coincidence.

The certification requirements for this type of aircraft do not address the failure of multiple systems as experienced on this flight. Accordingly, an engine failure, coupled with a total hydraulic failure leading to a manual extension of the landing gear, will leave the aircraft with high drag profile and low power that would be beyond the experience of most pilots operating this class of aircraft. The aircraft flight manual did not address this problem, nor was it required to, because the event is well outside the certification requirements, and it is unlikely that flight testing was carried out to establish performance parameters that would lead to flight manual recommendations.

5. Bureau determination

Based on the data made available to the investigation, the Bureau was unable to positively determine what initiated the failure of this engine. The Bureau is not convinced that an air foil separation due to non-discernible grain areas was the primary cause. However, if that was the primary cause, and if the manufacturer's service bulletin is properly addressing the removal from service of discrepant blades, then the incidence of engine failures due to this factor should rapidly decrease.

It should be noted however, that the one blade from this engine that was identified as having a non-discernible grain area, was from a series that were not listed for removal by the service bulletin.

If the failure was due to second stage low pressure turbine stator outer casing fatigue, then it is to be expected that more failures of this type could occur, with more definite evidence of fatigue able to be identified, resulting in a fleet rectification program.

VH-AWL was operating over the central business district (CBD) at 2000 feet when VH-AGS called Essendon tower at Point Ormond (PTO) requesting clearance also to operate over the CBD at 2000 feet. Due to workload the clearance was denied and VH-AGS was asked to sight VH-AWL. Later, a radar return was observed entering the control zone north of Point Ormond flying at the same level and in the opposite direction to VH-AWL. The pilot of VH-AGS was asked to sight VH-AWL which he did at a range of about one mile. There was insufficient time to alert VH-AWL. VH-AGS was about 1.5 miles inside the control zone at the time of the confliction.

The pilot of VH-AGS later claimed that he thought he was cleared to enter the control zone once he had sighted the other aircraft. He also thought he had been told the other aircraft was at 3000 feet. He had the other aircraft in sight from the time he entered the control zone until the confliction occurred. He estimated the two aircraft passed with 200 feet of vertical separation and about 200 metres of lateral separation.

The aircraft had completed four uneventful take-offs and landings. A short field take-off was then carried out. The landing gear was retracted and a procedure turn executed at 500 feet. On completion of the turn, landing flap was selected and the prelanding checks initiated.

The pilot reported that he became preoccupied with flying the aircraft accurately and overlooked selecting and checking that the landing gear was down. This was not realised until the throttle was reduced to idle on entering the flare. The aircraft settled onto the runway with the landing gear retracted and the gear warning sounding.

The aircraft was departing runway 16R at Sydney on a Radar 5 departure, with a clearance instruction to turn right heading 170 degrees after take-off. On becoming airborne, the aircraft was observed to turn left onto a heading of about 125 degrees. The pilot was asked to confirm heading to which he replied, "170". The aircraft was subsequently seen to turn right, onto the correct heading.

The take-off had been flown by the first officer, with the pilot-in-command providing support functions. The pilot-in-command was resetting an incorrect radio frequency selection as the aircraft made its initial climb and initially failed to notice that the aircraft was turning left instead of right. At the time ATC contacted the aircraft the pilot-in-command was instructing the handling pilot to turn right, to comply with the clearance. There was no conflicting traffic at the time of the occurrence.

VH-JCH was cleared by Albury Tower on the direct track from Albury to Canberra at 9000 feet. VH-CSV was on the same track flying from Canberra to Albury and believed (by Albury Tower) to be at 8000 feet. Accordingly, the pilot of VH-JCH was given a requirement to reach 9000 feet in time to avoid a conflict. This was complied with.

The pilot of VH-CSV called Albury Tower at 45 miles from Albury (10 miles prior to controlled airspace) and was cleared to Albury and to descend to 7100 feet which is the applicable route lowest safe altitude. VH-CSV then reported leaving 9000 feet. This was the first the Albury Tower controller knew that VH-CSV was at 9000 feet. The pilot was instructed to descend immediately to 8000 feet. This altitude change was probably achieved before the aircraft entered the control area and hence a loss of separation, technically, probably did not occur.

VH-CSV had originally maintained an 8000 foot cruise altitude but had later climbed to 9000 feet and advised Melbourne Flight Service of this change. The flight service officer was the only person on duty for this area of the network and a busy complicated period of work was in progress. She attempted to contact Albury Tower to pass advice of the level change but was unable to get through.

Albury Tower was also operating with one person on shift and was in a busy period. Uncontrolled airspace between Canberra and Albury between 8000 feet and 10000 feet spans a track distance of only 39 miles. Aircraft of the type involved in this incident transit that distance in a fairly short period of time. Due to the workload imposed by other tasks at both Albury Tower and Melbourne Flight Service, the co-ordination on the altitude change for VH-CSV was not achieved, although every effort was made.

Significant Factors

The following factors were considered relevant to the development of the incident:

1. Operation of air traffic service positions by one person during busy traffic periods.

2. Co-ordination on altitude change for VH-CSV, by the flight service officer, was not achieved.

The pilot of VH-LLY was cleared for descent in Darwin controlled airspace from FL330 to 5,000 ft, on a heading of 085, for radar vectoring to the ILS. VH-JPW had departed Darwin, and the pilot was cleared for climb to FL330, on a heading of 163.

During descent, as VH-LLY was passing through 8,000 ft with a 3,000 ft per minute rate of descent, the pilot heard Darwin Approach advise the pilot of VH-JPW of unidentified traffic in his 2 o'clock position, and to maintain 6,000 ft until clear. He looked out and sighted another aircraft rapidly approaching and realised that he must be the unidentified traffic. He resumed control from the co-pilot, made an immediate left turn and passed behind the other aircraft, with an estimated separation distance of 200m.

He then contacted the pilot of VH-JPW and Darwin Approach to advise them of the occurrence. The pilot of VH-JPW believed that if he had not maintained climb power when he levelled at 6,000ft, the passing distance would have probably been much less.

The Darwin Approach Controller admitted to misidentifying VH-LLY, contributing to the breakdown in separation standard.

A landing approach was made for runway 24 with the gear selected down and the cockpit gear lights indicating green. As the aircraft slowed on the landing roll the left main gear green light went out and the gear unsafe light came on. The aircraft was gently slowed and a left turn started when the left main gear leg collapsed.

Inspection of the main gear leg showed that the pin securing the top of the leg to the wing structure, had seized and then sheared. The pin then moved forward and allowed the top of the leg to detach from the wing support structure, and collapse.

Significant Factor

The following factor was considered relevant to the development of the accident:

1. The pin securing the top of the left main landing gear leg to the supporting structure seized and fractured. (The reason for the seizure was not determined.)

VH-EJM was being operated on dual circuit training on runway 30. A mixture of left and right circuits had previously been flown, with the last one being a right circuit. Stop and go landings were being made. On the last approach air traffic control had instructed the pilot to make the next circuit a left circuit.

VH-FNE was lined up on runway 35 and after VH-EJM had crossed the runway intersection the controller cleared VH-FNE for take-off. On climb the crew of VH-FNE observed VH-EJM to be approaching their flight path, on a right circuit. The climb of VH-FNE was stopped and VH-EJM passed about 200 feet above and slightly behind VH-FNE. The pilot of VH-EJM had seen VH-FNE.

At about the same time as the pilots of the two aircraft saw each other the air traffic controller, who had been looking in another direction at a third aircraft, looked back and saw VH-EJM flying a right circuit and in conflict with VH-FNE. Traffic information was given to VH-FNE and the pilot of VH-EJM was advised of the mistake and to continue with the right circuit.

Significant Factor

The following factor was considered relevant to the development of the incident:

1. Misunderstanding of the instruction to make a left circuit, by the pilot of VH-EJM.

The pilot reported that the aircraft accelerated normally during the take-off run and began a normal climb over tree covered rising ground. At about 100 feet above the tree tops a strong down draught was encountered, which the aircraft was unable to out climb. When it became evident that a collision was inevitable the pilot attempted to stall the aircraft into the trees. The aircraft collided with the treetops and nosed over, coming to rest inverted. Both occupants were able to vacate the aircraft with minor injuries.

Circumstances

The B737 was operating on airway B580 and had passed waypoint URGUS at 1234 EST maintaining FL350 and was estimating waypoint MULID at 1322. Flight plan estimates indicated that the B737 would cross airway A464 at approximately 1343 and the next waypoint APORA at 1424.

The B74F was operating on airway A464 and had passed waypoint RELIK at 1251 maintaining FL350 and was estimating over Lord Howe Island at 1323. Flight plan estimates indicated that the B74F would cross airway B580 at approximately 1331 and the next waypoint UDIKO at 1401.

At approximately 1254, Brisbane oceanic control sector 8 (Sect 8) relayed a level change requirement instruction through Sydney Flight Service International (FS INT) for the B737 to descend from FL350 to FL330. The instruction was intended to provide separation assurance between the B737 and the B747F.

The level change requirement was notated on the correct flight progress strip but entered in an incorrect information box relating to waypoint URGUS, the waypoint just passed.  The FS INT officer was provided with a temporary relief break at approximately 1310.  The level change requirement had still neither been passed to the B737 nor been included in the handover/takeover to the relieving officer.

At approximately 1328, the relieving FS INT provided Sect 8 with reports for the B737 passing MULID and the B74F passing Lord Howe Island. Both aircraft were still maintaining FL350. Sector 8 realised that the aircraft level change instruction had not been issued and that lateral separation standards had been infringed.

Due to the prevailing poor communications propagation conditions, FS INT was unable to re-establish contact with the B737 until approximately 1336. The B737 then descended to FL330 and crossed airway A464 approximately seven minutes later.

Although the prescribed separation standards were not maintained, there was no risk of collision.

No new safety deficiencies were identified during this investigation.