The pilot was conducting a flight from Diamond Beach to Cooma. Approaching Hoskinstown the aircraft's rear engine suffered a loss of power and smoke appeared in the cabin. The pilot immediately shut down the rear engine and feathered the propeller. Because the symptoms indicated that a fire was burning in the rear engine compartment, the pilot decided to land as soon as possible, and prepared for an emergency landing into a paddock. During the approach the smoke disappeared from the cabin, so the pilot concluded that the fire had extinguished itself and discontinued the landing. He then diverted to Canberra, using the front engine only, where he landed safely. An examination of the rear engine compartment revealed that a large section of the left-hand exhaust stack had separated in flight. Hot exhaust gases, now directed into the rear engine compartment, burnt a hole through the lower cowl, distorted the firewall and burnt the insulation from electrical wires including the low tension ignition leads. The bare ignition leads then "shorted out" causing the engine to fail. The source of heat having been removed prevented the fire from becoming self-sustaining. Cracks of about 18 cm in length had occurred along either side of the lower welded seam from the muffler attachment flange. The forward crack propagated around the exhaust stack, through 180 degrees, and returned back to the muffler attachment flange, paralleling the lower seam. This allowed the section of exhaust stack to separate. Expert opinion suggests that the cracks should have been evident for a considerable period of time, having originated from normal thermal fatigue and exhaust gas flow erosion. The aircraft had been subject to a major inspection a short time prior to the flight. The organisation which performed this inspection advised that there had been no evidence of exhaust gas leakage, or cracks visible to the naked eye at that time.

The pilot reported "Leaving Flight Level 370 for Flight Level 350" when the aircraft was about 320 kilometres to the north of Perth. The aircraft had not been cleared by Air Traffic Control to leave Flight Level 370 on descent. Investigation indicated that the pilot was only requesting a clearance to descend and his use of non standard radio transmission terminology gave the impression that the aircraft had already commenced descent. The aircraft remained at Flight Level 370 until cleared to descend by Air Traffic Control. There was no DEPARTURE from standard operational practice. This occurrence was not considered to be an air safety incident.

The aircraft was engaged in operating a night freight flight from Brisbane to Gladstone and Rockhampton. The pilot departed Rockhampton as a passenger in another company aircraft at about 2100 hours the night before the accident and was flown to Caloundra. He then flew VH-DIL to Brisbane and after arriving submitted a flight plan for the return flight to Rockhampton giving an estimated time of DEPARTURE Brisbane of 0230 hours. Before DEPARTURE, however, the pilot was required to await the arrival of another aircraft so that freight could be transhipped to his aircraft. DEPARTURE from Brisbane was made at 0301 hours and at 0400 hours the pilot reported the aircraft's position over Gayndah, the pilot was instructed to call on another radio frequency at 0410 hours. No further transmissions were heard from the aircraft. The wreckage was located two days later in mountainous terrain. The aircraft had impacted the ground in a near vertical attitude. Subsequent investigation did not reveal any fault with the aircraft or its systems that could have contributed to the accident. The day before the accident the pilot rose at about 0730 hours after spending the previous night and day at a friends property near Rockhampton. He spent the day at leisure at the property before returning to Rockhampton in the afternoon. After dining at his parents home he proceeded to the airport for the flight to Caloundra. Before departing Brisbane to return to Rockhampton the only sleep the pilot would have had was about an hour on the flight to Caloundra and possibly another short period at Brisbane Airport while waiting for his aircraft to be loaded.

The pilot returned to Essendon at about 1700 hrs ESuT on 4 December after having operated for some time in Tasmania. On his arrival he refuelled the aircraft and completed a flight plan for the flight to Adelaide and Kalgoorlie the following morning. At about 0430 hours ESuT the next morning the pilot submitted the flight plan to Melbourne Briefing Office prior to obtaining a forecast of the upper winds for the route and terminal area forecasts for Adelaide and Kalgoorlie. To calculate the time intervals for the two legs of the proposed flight the pilot had assumed a headwind component of 50 Knots. However, the forecast given to the pilot indicated that the headwind component for the Essendon to Adelaide leg at the planned level, flight level 310, was about 90 knots, and for the Adelaide to Kalgoorlie leg at planned level, flight level 370, was about 100 knots. The flight plan also gave the fuel endurance of the aircraft as 200 minutes ex Essendon and 300 minutes ex Adelaide. The aircraft subsequently departed Essendon and arrived in Adelaide after an apparently uneventful flight. The pilot left a note to instruct the aircraft refuellers to refuel the aircraft to full tanks and proceeded to the Adelaide Briefing Office. At the Briefing Office the pilot obtained updated weather forecasts which indicated that Kalgoorlie required an alternate. The pilot then advised the Briefing Officer that he would nominate Perth as the alternate. When he was then told that the fuel endurance necessary to plan Perth as an alternate would be 302 minutes, the pilot changed the flight plan to indicate that the fuel endurance of the aircraft ex Adelaide was 320 minutes. At the flight levels planned a fuel endurance of 320 minutes could not be obtained. The pilot also decided to fly the leg at flight level 290 as the head winds at that level were less than at the higher level, although still almost twice the strength of the flight planned headwind. When the pilot returned to the aircraft he found that it had not been refuelled as requested. He then assisted the refuellers to refuel the aircraft by adding the anti-icing agent to the fuel during the operation. Because of the delay the normal fuelling procedure was not carried out and as a result the fuel tanks were not filled to capacity. About an hour after departing Adelaide the aircraft entered cloud and encountered icing conditions. The pilot switched on the engine anti-ice and the cruise was continued at flight level 290 for about 30 minutes before climbing to flight level 310, clear of cloud where anti-icing was turned off. Operating with engine anti-ice on increases the fuel consumption by 8 percent. As the aircraft approached the mid-point of the flight the pilot became concerned that he may not have sufficient fuel to continue the flight from Kalgoorlie to the alternate, Perth. He decided to continue the flight towards Kalgoorlie and if the alternate requirement was not lifted when he was at a point along track that was 30 minutes beyond Caiguna, he would return to and land at Caiguna. Shortly afterwards the alternate requirement was lifted on Kalgoorlie. When 185 kilometres east of Kalgoorlie the pilot commenced the descent and at flight level 250 the aircraft entered cloud and the engine anti-ice was again switched on. During the descent the fuel low level warning light illuminated and the left engine surged and flamed out at 10,000 feet. The pilot unsuccessfully attempted to restart the engine. As the aircraft broke clear of cloud at 1000 feet above ground level the right engine also flamed out. The pilot made a distress call and landed the aircraft, gear up, on a fire break. After touchdown the aircraft skidded 400 metres before coming to rest. The examination of the aircraft found that during the landing the left fuel tank had been breached, however no evidence was found of any significant fuel spillage. Approximately five litres of fuel was recovered from each of the left and right fuel tanks. The engine fuel filters and fuel lines provided only a small amount of residual fuel. The inspection of the remainder of the aircraft did not reveal any defects that could have contributed to the accident.

The pilot had been practising circuits and landings in light and variable wind conditions using the 170 degree grass strip. He made an approach for a practice short field landing with full flaps selected at an approach speed of 38 knots (stall speed in that configuration is 25 knots). The pilot said that as he approached the ground the rate of descent was high so he let the airspeed increase to 40 knots to initiate the flare. Rate of descent was still too high so he applied power. The aircraft landed in the three point attitude. The pilot described the landing as heavier than he would have liked. The aircraft did not bounce but ran straight and then started to sag on the right side. The right main gear diagonal brace had broken and the remainder of the structure gradually bent under the weight of the aircraft. The pilot steered the aircraft off the grass to the right and stopped. The broken diagonal brace was inspected and was found to have failed due to overload. The pilot advised that the main landing gear bungees were not the normal ones used on the Auster. They were in fact of a type that is fitted to the Piper Pawnee. The reason for this was that the correct ones for the Auster were not available at one stage so it was common practice to use the type fitted to the Pawnee. The effect was that the main landing gear was much stiffer. This was considered a possible factor in the failure of the diagonal brace. It was also determined that the pilot had not made any short field landings for some time. It was for that reason that he decided to practice a short field landing. This accident was not the subject of an on-scene investigation.

At about 1000 hours local time on 19 June 1985, Garuda Boeing 747 aircraft PK-GSB, operating as Garuda Flight 898, was being radar vectored by Melbourne Air Traffic Control (ATC) for a landing on runway 34 at Tullamarine Airport. Shortly after the pilot acknowledged an instruction to call Melbourne Tower the aircraft was seen to turn and descend as though making an approach to runway 35 at Essendon Airport. The Melbourne Tower controller instructed the aircraft to climb to 3000 feet above mean sea level (amsl), and a few seconds later the aircraft was seen to be established in this climb. After further radar vectors were given,an uneventful landing was carried out at Melbourne. It was subsequently determined that the aircraft had descended to a minimum height of approximately 350 feet above the level of Essendon Airport, and was about 1.5 kilometres from the runway threshold before the climb was commenced.

Circumstances:

The aircraft was being pushed back from the parking bay of the operating company. The despatch engineer, who was on headset to the flight deck, was located near the nosewheels prior to commencement of pushback. When pushback commenced, he walked on the right side (looking to the rear of the aircraft) of the nosewheels, slightly leading them in the direction of travel and approximately one metre from them. Shortly after pushback commenced, the tug driver initiated a turn such that the nosewheels turned towards the despatch engineer. At this point, the despatch engineer was seen to stumble and fall into the path of the oncoming nosewheels, one of which ran over his right lower leg. The tug driver immediately applied the brakes, selected reverse gear, and pulled the aircraft off the engineer's leg. The investigation determined that the injured engineer had been in the employ of the company for one year at the time of the accident. He had done approximately six months on-the-job training in tarmac duties including pushback procedures. He was considered competent to do pushbacks of all aircraft in the fleet. The investigation did not determine what caused the engineer to stumble and fall. Company procedures for receipt and despatch of aircraft are published. Included in those procedures is a warning that "the despatch engineer must remain on one side of the aircraft, clear of nosewheels, whilst the aircraft is moving". The procedures also include advice that "the aircraft will normally be pushed back to a nominated despatch area with the despatch engineer walking beside the tractor and adjacent to the driver". The despatch engineer was aware of these published procedures. The reason he did not follow the procedures was not determined.

Significant Factors:

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

1. The despatch engineer was too close to the nosewheels during pushback.

2. The despatch engineer stumbled and fell into the path of the nosewheels resulting in his right lower leg being crushed.

Recommendations:

The major part of this investigation was done by the Safety Department of the operating company. As a result of that investigation the company made a number of recommendations which are summarised as follows

1. Issue instructions to ensure that all personnel involved in despatch of aircraft keep at least three metres away from the nosewheels and always stay behind the direction of travel of the wheels. In addition, the person on headset should stay on the inside of any turn.

2. Investigate the feasibility of a small hook on the inside of the Boeing 767 nose gear doors to stop the headset cord from draping over the nosewheels.

3. Evaluate cordless communication headsets for use by the despatch engineer.

4. Initiate a positive practical training program which incorporates a recording system to show the level of competence achieved by each employee.

5. Establish a reporting system with the Federal Airports Corporation which will enable personnel to report all damage to tarmacs that could constitute a tripping hazard.

6. Recommend to the Federal Airports Corporation that all tarmac earthing points be flush mounted to eliminate them as tripping hazards. In this investigation it was not determined what caused the engineer to trip but the consequences of tripping and falling were tragically demonstrated. Accordingly, part of the investigation effort focussed on potential tripping hazards which is the reason for some of the above recommendations. In addition, the company's Safety Department believed that procedures concerned with training engineers for receipt and despatch of aircraft should be more tightly controlled and better records kept. The Bureau of Air Safety Investigation supports the recommendations made by the company and further recommends that other companies with engineers involved in pushback of aircraft consider this report and recommendations in relation to their own operations.

The helicopter had operated normally during the first flight of the day. After DEPARTURE on the second flight the aircraft overflew Pyrmont before heading east along Sydney Harbour. As it approached the Sydney Heads, at 500 feet above sea level, the engine failed. The pilot gave a mayday call and successfully ditched the helicopter some 500 metres from two Army landing craft. After landing in the water the helicopter rolled inverted and floated for about 10 minutes. The crew escaped from the aircraft and after spending five minutes in the water were picked up by the landing craft. Prior to leaving the water the pilot attached a line from the landing craft to the helicopter. The helicopter subsequently sank but remained attached to the line and was towed to Balmoral Beach. An inspection of the wreckage revealed that a material failure of the acceleration bellows in the engine fuel control unit had caused the engine failure. The reason for the material failure was not established. The component has been forwarded to the manufacturer for further examination.

Circumstances:

At approximately 1809, (23 minutes before last light) during final approach to landing at Carnarvon, the pilot noticed that the landing gear had not extended correctly. The aircraft remained in the circuit area whilst the pilot attempted to lower the landing gear using both manual and emergency methods. He also sought assistance from the company's, Perth based, duty pilot and Carnarvon based engineers. After exhausting all possible methods of lowering the gear the pilot decided to land with the landing gear and flaps retracted. The pilot rejected a landing on the sealed runways because he was apprehensive that it would cause unnecessary damage to the aircraft and could result in a fire. He considered landing in a riverbed (rejected by the Senior Operational Controller), alongside one of the sealed runways (the surface was unsuitable) and on one of the dirt strips. The pilot was offered a flare path on dirt runway 27 however, he declined and indicated that he would try to land using the available light. At 1856 (last light was at 1832) the pilot attempted a landing on runway 27. On late final approach the aircraft collided with a one and a half metre high levy bank, 270 metres short and 115 metres to the right of the threshold. The pilot was trapped in the wreckage for some time after the aircraft came to a stop. The landing gear problem arose when the left main landing gear would not lower. Examination of the aircraft revealed that both hinges fitted to the inboard landing gear door had fractured. The forward hinge had fractured as a result of fatigue and the rear hinge as a result of overload. The fatigue crack initiation had occurred at a sharp edged, prominent forging flash on the inner radius of the hinge and had grown over approximately 4000 load cycles. A similar fatigue problem had been identified on an earlier version of the hinge (part number 46653-00), however, regular inspections for fatigue cracking were discontinued when hinges with part number 47529-32 (as fitted to VH-DEG) were introduced in 1980. Similar fatigue cracking was found in the forward door hinge of another PA31 during the investigation. The fractured hinges jammed the left main landing gear mechanism and neither the normal or emergency extension systems could extend the gear. The pilot was apprehensive about wheels up landings. Much of his decision making was aimed at reducing the risk of fire and minimising the damage the aircraft would sustain during the landing. e.g. Selection of a dirt runway instead of the sealed strip, landing with flaps retracted etc. During the pilot's attempts to rectify the landing gear problem, and up until the time of his touchdown, he was subjected to considerable radio transmission traffic involving questions, directions and suggestions which distracted him from his primary tasks. The pilot indicated on at least two occasions that he was ready to land, however, each time advice and questions from the ground personnel involved overrode his intentions. When the pilot was asked if he wanted a flare path on runway 27 there was still some natural light available and he was intending to land. However, by the time he was able to make his final approach it was dark and he was unable to see the ground. Studies have shown that aircrew subjected to high levels of stress can suffer skill fatigue and cognitive task saturation, which in turn can lead to a breakdown in the decision-making process. It was apparent from the pilot's radio transmissions and the quality of the decisions made in the latter part of the flight that his information processing and decision-making abilities had been degraded by the stress of continuous radio transmissions and continuous, and sometimes conflicting, instructions. As a result, what should have been a relatively simple wheels up landing in daylight was turned into an extremely difficult wheels up landing at night. With the landing gear retracted the aircraft's taxi and landing lights were not available to the pilot.

Significant Factors:

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

1. Manufacturing defect. A forging flash created a stress concentration which led to fatigue cracking.

2. Inadequate inspection procedures. Previous inspection procedures introduced to disclose similar cracking were withdrawn on the introduction of later part numbered hinges.

3. Apprehension of the pilot. The pilot was apprehensive about apparently significant dangers of landing an aircraft, wheels up, on a sealed runway.

4. Inordinate interference in aircraft operations by ground-based advisors. The ground advisors input overrode the pilot's decision on a number of occasions with the result that a simple exercise became very complicated.

5. Cognitive task saturation and skill fatigue. The amount of information, advice and suggestions being passed via the radio communications system overloaded the pilot decision making abilities.

6. Improper in-flight decisions. As a result of task saturation, the final decision made by the pilot to attempt a night landing on an unlighted strip was incorrect.

7. The pilot did not see and therefore was unable to avoid the levy bank.

Recommendations:

That the Civil Aviation Authority give consideration to amending AD/PA-31/93, to include the later part number hinges (P/N 47529-32 and 46653-00 (steel) which were introduced by SB 682) for a repetitive inspection in addition to the inspection called up for Pre SB 682 hinges.

Earlier on the day of the accident the pilot had experienced a loss of engine power during an attempted takeoff. The problem did not reoccur during a static engine run-up, and the pilot assessed the possible cause as being the hot air temperature. Flight was again attempted later in the afternoon, however, shortly after takeoff at approximately 60 feet above ground level the engine ran roughly, lost all power and produced black smoke. The pilot closed the throttle and attempted to land straight ahead. The aircraft touched down at the end of the strip, entered an overrun area of soft sand and overturned. Examination of the engine revealed that it had been operating on an excessively rich mixture. It was found that an incorrect model carburettor had been fitted, which would have provided too rich a mixture. The engine was being operated using a mixture of mogas and avgas, and the synthetic float in the carburettor was found to be significantly heavier than that specified due to absorption of mogas products. Tests indicated that the resulting high level of fuel within the carburettor float chamber, when combined with pitch changes associated with takeoff and initial climb, caused significant power loss and rough running due to spillage of fuel into the carburettor throat. Simulating the effects of turbulence and vibration exacerbated the power loss, and it is probable that the engine had lost all power due to a rich cut.