Prior to DEPARTURE the pilot had received a weather forecast which indicated that fog could be expected at the destination. Adequate fuel was carried in the aircraft tanks to allow for a considerable period of holding and/or a diversion to another aerodrome if required. After an apparently normal flight of 35 minutes the aircraft arrived in the Goulburn area and the pilot reported his intention to carry out a standard instrument approach. The manoeuvres associated with this procedure do not involve flight over the city of Goulburn, however witnesses observed the aircraft as it circled over the city several times at a relatively low height. It was then seen to roll and descend steeply before striking two houses. A fierce fire broke out which engulfed the aircraft and both residences. The three persons on board the aircraft and a person in one of the houses received fatal injuries. A detailed inspection of the wreckage revealed that the camshaft of the left engine had failed in flight and the pilot had apparantly feathered the propeller. The engine was not operating at the time of impact. No other defect or malfunction was discovered which might have contributed to the development of the accident. It was determined that the particular camshaft had failed from fatigue cracking, resulting from defective manufacture. It was considered likely that when the aircraft arrived over Goulburn, shallow fog obscured all or part of the aerodrome. The lights of the city would have been clearly visible and the pilot probably decided to use the city, rather than the nearby radio navigation aid, as a convenient holding point while waiting for conditions at the aerodrome to improve. During a series of left hand orbits, and after advising his intention to conduct an instrument approach, the pilot experienced a complete failure of the left engine. In order to realise the available single-engine performance of the aircraft the pilot had to perform a series of checks and actions which would result in the applicable propeller being feathered; any unnecessary aerodynamic drag being reduced; and an appropriate airspeed being established. It was determined that although the propeller had probably been feathered, the landing gear, which had evidently been lowered previously, had not been raised to reduce drag. In addition, an analysis of radar returns from the aircraft, recorded at Canberra, indicated that the airspeed at which the aircraft was flying shortly before radar contact was lost, was less than the optimum figure. The final manoeuvre described by witnesses was consistent with that which follows a loss of control in twin engine aircraft when power is being supplied by only one engine and the speed is below the minimum required for full control. The reason the pilot did not raise the landing gear and maintain the required minimum control speed could not be established.

Circumstances:

The pilot was conducting a travel flight over two days, from Lilydale to Essendon, Merimbula, Mallacoota and returning to Lilydale. Pre- DEPARTURE, he checked the fuel tanks and established they were full. The left tank gauge was indicating full, while the right tank gauge was showing less than half. On setting course for Essendon, he noticed the cylinder head temperature gauge was indicating a high temperature so decided he would use a cruise mixture setting richer than normal. He selected a fuel flow of 105 pounds an hour (68.6 litres per hour.) On the subsequent legs the pilot adopted a similar procedure. Before departing Merimbula the next morning, he checked the tank levels by putting his finger in the tank aperture and ascertaining the fuel was about 7.5 centimetres below the top. He decided there was adequate fuel for the rest of the flight and did not refuel. On the flight plan he recorded an endurance of 240 minutes, with a plan time for the two legs of 110 minutes. The flights proceeded normally, and the pilot again used the higher cruise fuel flow. After passing the East Sale area and when heading towards Warragul he noted the right tank fuel gauge was on empty and the left was on one quarter. He decided that the time flown, and his preflight determination of tank contents meant there was adequate fuel for the rest of the flight and continued. About 10 minutes out of Lilydale the engine failed, apparently due to fuel exhaustion in the left tank. The right tank was selected and the engine regained power. The aircraft approached Lilydale for an entry on the dead side of the circuit. While at circuit height for a landing into the north, the engine failed. The pilot manoeuvred the aircraft via an "S" turn onto a final approach to the right hand northern runway. The wind was from the north at 15-20 knots and on turning onto final the pilot realised the aircraft would not make the strip. The aircraft contacted the ground 50 metres short of the airfield boundary with the landing gear up. The VDO time meter recorded a total engine running time of 4.9 hours since departing Lilydale. There was no fuel left in the tanks. In his flight plan considerations for endurance the pilot did not make any allowance for the extra fuel required for each take-off and climb to cruise altitude. In addition, he did not revise his fuel endurance to take account of the use of a higher than planned cruise fuel flow. Calculations made during the investigation showed that when these factors were taken into account, planned fuel usage was consistent with actual fuel used.

Significant Factors:

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

1. The pilot's fuel planning did not take account of the extra fuel required for several take offs and climbs.

2. The pilot used a different fuel flow in flight to that which he had used for calculating the flight plan endurance, but he did not recalculate endurance.

3. The pilot continued the flight after he had observed fuel gauge indications that were abnormal.

4. Engine failure due to fuel exhaustion.

5. The pilot misjudged the forced landing approach.

The pilot was conducting a sight-seeing flight and was returning to Jabiru at 6-700 ft above ground level. The pilot recalled seeing a bird just before it impacted the top of the windscreen. The bird did not enter the cockpit but the perspex was shattered. A large piece of windscreen attached to the compass electrical wires swung inwards and hit the pilot across the forehead, nose and mouth. He was wearing sunglasses at the time which may have protected his eyes from the debris. The pilot was able to control the aircraft normally although extra power was required to overcome the increased drag. The aircraft returned safely to the DEPARTURE airfield.

There had been rain on the strip and the surface was damp with muddy patches. A witness observed the aircraft become airborne after one half to three quarters of the strip length and climb to 5-7 metres above ground level. The witness looked away at this point but when he checked a short time later the aircraft had disappeared. The wreckage was located some 300 metres beyond the end of the strip in a tidal mangrove swamp. The pilot, who had limited experience in bush operations, reported that the first indication he had that all was not normal was when he crossed the upwind end of the strip at a height lower than expected. From this point on, the aircraft slowly descended into the mangroves. He did not detect any abnormality with the aircraft or its engines before or during the takeoff. An examination of the engines revealed that one and possibly both magnetos on the rear engine were not functioning correctly at impact. This would have reduced the aircraft's performance in the manner described by the pilot. The condition of the strip, coupled with the pilot's low experience level, could have reduced the prospect of him detecting a loss of performance much earlier than he did whether the magneto failure(s) occurred before or after liftoff. This occurrence was not the subject of an on-site investigation.

When the landing gear was lowered in the circuit normal down and locked indications were obtained. During the landing roll the right main gear collapsed. Investigation revealed that a damaged "O" ring in the right downlock actuator had allowed hydraulic pressure to be lost, causing the right gear to unlock and collapse. The actuators are "on-condition" items.

The pilot had shut down the helicopter to kill and load meat for a stock camp. After starting and completing the after-start checks, the pilot first entered a low hover. The aircraft was then put into a high hover about 10 ft above the tree tops before a climb was initiated into a wind of 15-20 kts. The pilot reported that as he accelerated through about 40 kts, he noticed a substantial reduction in the manifold pressure reading with the engine RPM at about 3000. The aircraft then began to sink into the tree tops. Some 100 metres further on, the main rotor blades struck a large tree trunk and the aircraft fell to the ground on its right side. The pilot was rendered unconscious at impact, but after regaining consciousness he assisted the surviving passenger from the wreckage and freed the other passenger's leg from under the cabin before dragging him clear. A strip inspection of the engine and turbocharger revealed significant defects. These included the density controller was set below tolerance; a cracked and warped wastegate butterfly; and a sticking wastegate actuator. Physical damage to the engine was attributed to poor engine handling. The cumulative effects of these defects would have reduced the available power output from the engine, but do not in themselves explain the symptoms reported by the pilot. In addition, the power available was sufficient for the pilot to execute a towering take-off. A turbocharger failure would have produced the loss of power experienced by the pilot, but no possible cause or result of such a failure was discovered. The pilot reportedly did not make a significant reduction in power after the high hover, so that it is unlikely that a sticking wastegate caused the problem. The cause of the reported loss of power was not determined.

The helicopter approached the "London Bridge" feature from the west along the coastline at about 400 feet above sea level. It then carried out a wide, right orbit of the feature and descended in an easterly direction as if to position the helicopter for the passenger to photograph "London Bridge" or perhaps to land in the car park. While facing east and still descending through an estimated 200 feet, almost above the feature, at slow forward airspeed, the helicopter began turning to the right without banking. The rate of turn accelerated rapidly. At about the same time an increase in engine noise was heard by witnesses who soon saw the aft section of the tail boom, encompassing the tail rotor assembly and stabilisers, separate from the helicopter. With the tail boom off, the helicopter adopted a steep nose down attitude as it rotated and plummeted. Witnesses estimate that the helicopter rotated through about six complete turns before it disappeared over the cliff edge. At the same time pieces were seen falling from the helicopter. In the soft rock on the lip of the cliff, about 65 feet above the sea, three distinctive rotor strikes were found. Two of the four main rotor blades were torn off at the blade grips probably when the blade tips impacted the cliff. These two blades were lost in the sea. At least one rotor blade hit the upper left rear of the fuselage, probably after striking the cliff. The air intake area, the structural members above the engine bay and the tail boom sustained massive impact damage consistent with main rotor strike(s). The outboard one third of the fourth blade appeared to have been broken off when the helicopter finally impacted hard in shallow, surging seas a few metres west of the cliff base. About three metres of tail boom was also lost in the sea. The fuselage, the rotor systems, the engine, the gearboxes, and the drive-train were examined by engineers. All damage was consistent with overload failure. Except for a small portion of the left rear door and remains of hinge assemblies, all four cabin doors were lost in the sea. The left rear door lower hinge pin was not found, nor was any remnant of the lower hinge bracket. It is unlikely that the hinge pin came out during the accident unless it had been inserted improperly. The only other alternative is that the pin was missing. The hinge pin incorporates a spring loaded bead near its tapered end and a safety tang at the top to prevent it working upwards and out. No structural damage was found in the lower hinge cutout in the fuselage. In contrast, the upper hinge bracket on the left rear door failed through overload, consistent with the door having been thrust upwards about the upper hinge without restraint from the lower hinge, while the door was in the open position. A metallurgical study found that the upper hinge had also previously sustained a partial fracture which was probably present for a long time. The tail rotor blades showed evidence of having sustained impact damage in flight. The only portion of the door frame which was found displayed a dent as if it had been hit by a rotor blade. An indentation on the leading edge of one of the tail rotor blades was consistent with it having hit part of the door latching mechanism. Witnesses saw the tail rotor blades at about the same time that the helicopter began rotating to the right. This would suggest that the tail rotor blades had already slowed down considerably because at normal operating RPM the tail rotor is seen as a blurred disc. It is probable that the left rear door came off during the low speed descent and struck the main rotor blades which spin anticlockwise when viewed from above the helicopter. The door, or fragments of it, were probably flung into the tail boom and into the spinning tail rotor. The tail rotor drive shaft broke from overload, probably at the moment the door struck the tail rotor. Then the tail rotor would have slowed down rapidly due to aerodynamic drag plus the stub of drive shaft flailing within the tail boom. The fuselage would have begun rotating to the right because there was no longer any tail rotor thrust to counteract engine torque. Witnesses heard an increase in engine RPM, probably at about the same time as the anti-torque failure. This is consistent with a sudden removal of the engine load needed to drive the tail rotor especially during low speed flight. The aft section of the tail boom showed evidence of having been struck by a main rotor in flight. This could have come about by the affects of the door impacting the main rotors and disturbing their tip path plane or by the pilot losing control during the anti-torque failure. Once the tail boom separated, the change in the helicopter's centre of gravity would have been outside fore and aft cyclic control limits and the pilot would have lost control. After identifying the anti-torque failure, it is probable that the pilot instigated the rapid descent by entering autorotation in an attempt to carry out the emergency drill. A subsequent experiment with the slam shut doors, as were fitted to this helicopter, found that the rear doors could be slammed shut with minor difficulty despite the absence of the lower hinge pin. With all the latches properly adjusted, the door might remain shut in flight but its alignment may be slightly askew and the door would be more prone to open in flight. About twenty-five minutes before the accident, witnesses saw the left rear door open during the hover/taxiing phase at Warrnambool. The pilot landed the helicopter and shut the door before proceeding to "London Bridge".

Circumstances:

The aircraft was cruising at 1500 feet above ground level when the engine lost power. As the engine stopped there was a cloud of white smoke and a loud noise. The pilot lowered the collective control and as the aircraft entered autorotation, he turned through 180 degrees to make an approach into the only reasonable area. The surface conditions were rougher than the pilot had expected and whilst attempting to reduce forward speed for the landing he misjudged the altitude, and the tail struck the ground. The pilot had not completed an autorotational approach and landing for approximately two years although he had completed an autorotational approach, with power recovery prior to landing, about four months before the accident. The engine failure was caused by a fractured crankshaft. The fracture followed the growth of a fatigue crack in the No. 4 connecting rod journal. It was apparent that the bearing shells fitted to the big end of the No. 4 connecting rod had rotated in their housing and had moved forward to contact the forward fillet radius of the No. 4 connecting rod journal. Fatigue crack initiation occurred as a result of the damage caused by the contact of the rotating bearing shells. The fatigue cracking had occurred during the fourteen-hour period of operation since the last overhaul. Fretting damage on the mating surfaces of the No. 4 connecting rod big end bearing housing indicates that the clamping force achieved by the connecting rod bolts had been poor thus allowing the bearing shells to move. Other deficiencies, in the way in which the engine was assembled and which were not connected with the engine failure, were found. Although some of the deficiencies may have been caused at the time of the crankshaft failure, others indicated that the engine may have been delivered to the operator, following the overhaul, in a defective condition.

Significant Factors:

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

1. It was possible that the engine overhaul had not been completed correctly. Indicated by the deficiencies found during the post-accident investigation.

2. There was a fatigue failure of the crankshaft causing complete engine failure.

3. The pilot misjudged the aircraft's altitude during the termination of the autorotational descent, and he was unable to prevent the aircraft's tail from striking the ground.

4. The landing area terrain was rough and uneven.

At about 50 knots during the landing roll, the pilot discovered that the right wheel brake had failed. The aircraft had swerved to the left when both brakes had been applied. With the rudder control ineffective at low speed, the pilot intentionally ground looped the aircraft to avoid running into the perimeter drainage ditch. The right main gear collapsed and the right wing was damaged as a result of the ground loop. Examination of the right brake unit revealed that the master cylinder was corroded internally, allowing brake fluid to bypass the seal.