While the aircraft was descending to land at Canberra the forward toilet smoke detector activated, the flight deck smoke annunciate light illuminated and a small amount of smoke was visible on the flight deck. The flight attendant confirmed there was no evidence of fire in the toilet. A slight haze was visible in the cabin with wisps emanating from the air-conditioning vents. The crew continued with the landing and a normal disembarkation ensued.

Oil was found to be dripping from the engine which was removed for rectification. The number 1 air/oil seal was found to have delaminated allowing engine oil to permeate into the air-conditioning ducts. The seal had previously been repaired at the manufacturer's facility. After this incident the operator determined to use only newly manufactured seals and introduced an in-situ pressure test of the seal.

As the aircraft was about to land at the property strip at night a car positioned at the end of the strip with its headlights shining along the strip.

During the landing the aircraft struck the car with the left main gear and crashed onto the runway with the gear collapsed. The pilot was aware that the car was parked near the end of the strip.

The pilot reported that the landing gear collapsed late in the landing roll. An inspection by a Licenced Aircraft Maintenance Engineer failed to find any pre-existing defects.

An accident diagram submitted by the pilot indicated that all the landing gear collapsed at about the same time.

A Boeing 747 aircraft was departing from runway 16R at Sydney whilst a DeHavilland DASH 8 aircraft was departing from runway 16L. As the Boeing 747 reached about 800 ft, the crew received a Traffic Collision and Avoidance System (TCAS) resolution advisory (RA) alert which advised them to monitor the aircraft's vertical speed.

The crew of the Boeing 747 were able to observe the DASH 8, airborne from the parallel runway, and ascertained that there was no danger of confliction. The TCAS alert was of very short duration. The Boeing 747 crew maintained their current rate of climb, monitored the TCAS and did not deviate laterally or vertically from their cleared route. The routes of the two aircraft were divergent and separation increased rapidly as the flight of both aircraft progressed.

A radar plot of the two aircraft confirmed the Boeing 747 crew's perception that there was no confliction with the other aircraft. The plot indicated an initial horizontal separation of 0.6 NM which reduced momentarily and then diverged in accordance with the cleared routes of both aircraft. Vertical separation of over 1,000 ft and lateral separation of 3 NM was achieved when the Boeing 747 was four miles from the aerodrome.

The radar plot showed that the DASH 8 became airborne just prior to the Boeing 747. Because the Boeing 747 was behind and had a significantly higher rate of climb in comparison to the DASH 8, it is likely that the TCAS assessed the DASH 8 as approaching the Boeing 747 and the system generated a preventative RA alert. Subsequently, as the Boeing 747 accelerated and passed abeam whilst climbing through the other aircraft's level, the TCAS reassessed the situation and cancelled the RA. There was no breakdown in separation.

It was reported that the pilot and passenger were conducting a stock survey at 200 ft above ground level when there was a loud noise and severe vibration from the area of the engine. The pilot immediately entered an autorotation descent and turned the helicopter into wind. A check indicated that no power was available from the engine. The terrain was rocky and covered in trees, so the pilot attempted to cushion the landing as much as possible. On touchdown one skid collided with a large rock causing the helicopter to roll over onto its side. It caught fire immediately and the passenger, who was the first to exit the wreck, assisted the pilot to escape. The helicopter, including the emergency locator beacon, was destroyed by fire.

The operator maintained its own search and rescue watch on its aircraft, and a search was commenced as soon as the aircraft was reported as overdue. The pilot and passenger were found the following day by search aircraft attracted by a signal fire lit by them.

Post accident inspection indicated that the engine problem was caused by the failure of a big-end bolt in the conrod of the no.1 cylinder which allowed the conrod to come loose and punch a hole in the crankcase. The engine had been overhauled by the manufacturer 300 hours (operating time in service) prior to the accident. The fire was probably started by oil that had leaked from the hole in the crankcase onto the exhaust system.

The pilot reported that soon after take-off at a height of 150 to 200ft the engine spluttered and then stopped. The aircraft was landed on the remaining runway. The left main landing gear was torn off and the right wing was damaged during the landing. The cause of the engine stoppage was apparently caused by fuel flow interruption, however extensive post flight examination could find no fault with the engine or fuel system.

VH-OJD was cleared to track inbound via the 229 VOR radial on descent to 6,000 ft. The Approach South radar control sector controller (APP S) was processing the aircraft for landing on runway 16R. Due to operational requirements, the APP S controller delayed turning the aircraft onto a right downwind for that runway.

VH-OLN had departed runway 16R, cleared initially to climb to 5,000 ft. The Departures South radar control sector controller (DEP S) cleared the aircraft for an early right turn and an unrestricted climb to its planned cruising level.

The APP S controller was about to instruct the crew of VH-OJD to turn onto right downwind for runway 16R when he recognised that VH-OLN was climbing through 5,000 ft. In an attempt to maintain lateral separation, he instructed the crew of VH-OJD to turn left immediately and advised them of the traffic ahead to their right. At this time, the crew of VH-OJD advised that they were complying with a TCAS Resolution Advisory and climbing to 7,000 ft.

Recorded radar information indicated that the aircraft had closed to approximately 2.1 NM, with 400 ft vertical separation when the minimum separation should have been 3 NM or 1,000 ft vertically. Each had penetrated their respective airspace boundary buffer.

The runway 16 airspace agreement has a narrow corridor in the APP S airspace to the west of the airport. This airspace design limits the flexibility available to APP S when vectoring aircraft to downwind for runway 16. The latitude that DEP S may apply when vectoring departing aircraft onto south-westerly headings, and providing climb, is also restricted.

Separation was not ensured before VH-OLN was cleared for an unrestricted climb. The situation was worsened by the high rate of climb achieved by VH-OLN.

Findings

1. The unrestricted climb instruction provided to VH-OLN by the DEP S controller was premature.
2. The vectoring of VH-OJD onto right downwind for runway 16R by the APP S controller was initiated late.
3. The airspace agreement boundary for runway 16 operations restricted flexibility in the processing of both arrival and departure aircraft.

Factors

1. The radar surveillance techniques used by both controllers were inadequate to prevent the loss of the prescribed separation standards.

Safety action

As a result of occurrence 9503423, the Bureau of Air Safety Investigation made interim recommendation IR950212. As the interim recommendation is also applicable to this investigation, it is reproduced below.

IR950212

The Bureau of Air Safety Investigation recommends that Airservices Australia amend the runway 16 airspace agreement boundaries to ensure that arriving and departing aircraft cannot be on, or near, reciprocal tracks. The amendment should address excursions into either buffer area.

The pilot reported that shortly after departure from Jandakot the engine lost power and then stopped altogether. He attempted to carry out a forced landing in a grassed paddock, but the aircraft was damaged in the process.

An inspection of the aircraft's fuel system disclosed that the fuel filter, lines and carburettor were contaminated by a significant amount of rust-coloured water. However, no water was evident in fuel obtained from the drain points on either the main or auxiliary fuel tanks. The pilot reported that he had drained a significant amount of water from the main fuel tank prior to departure.

The main fuel tank was constructed of corrugated metal. The longitudinal corrugations were not connected to each other, and most were not connected to the single drain-sump. Water had collected in the corrugations, and some was still evident after the accident. The drain-sump was not located on the lowest part of the tank and undrainable water could accumulate in the tank, aft of the sump, with the aircraft in the ground attitude.

An experienced Tiger Moth engineer reported that inadequate fuel tank drain systems was a known problem, and an Airworthiness Directive (AD) had been issued which required that fuel-drain points, fitted to some aircraft types (including the Tiger Moth), be designed so that water could not be conveyed to the engine in any normal attitude for the aircraft. These drain points were usually fitted to the lowest point on the tank. That is, along the rear edge. The AD was withdrawn when a number of requirements, covered by several ADs, were incorporated into Civil Aviation Orders Part 100. Transfer of the fuel-drain requirements to the CAOs was incomplete because, whilst similar fuel-drain requirements exist for aircraft constructed in accordance with the amateur-built category of aircraft there is no longer any requirement to retrofit similar systems to existing aircraft, such as the Tiger Moth.

Two Tiger Moth fuel tank modifications, designed to meet the requirements of the AD before it was withdrawn, were available to overcome the lack of drainage found in this tank. Both incorporated drain points along the rear edge. Neither of these modifications had been incorporated in VH-HPH's fuel tank. It is probable that in-flight aircraft movement allowed some of the water that had collected in the corrugations, and at the rear of the tank, to make its way into the fuel system and cause the engine stoppage.

An inspection of the ground marks at the accident site indicated that the aircraft touched down whilst it was still in a sideslip to the left. Touchdown was followed by a partial ground-loop as the aircraft tipped up onto its nose and left upper wing. The aircraft was approaching a fence on touchdown. It is possible that the pilot used the sideslip manoeuvre in an attempt to lose height quickly and touchdown earlier but did not stop the sideslip in time for the landing.

An inspection of the lower left wing during the post-accident investigation disclosed that one spar had fractured some time prior to the accident. It was also evident that reconstruction of the wing, also carried out some time prior to the accident, was not in accordance with the manufacturer's specifications.

A Squirrel AS 350B helicopter, and a Jetranger 206B helicopter, were conducting scenic flights in the Sydney area on the afternoon of the occurrence. These flights usually passed over Middle Harbour but, because runway 25 was in use, they were cleared to track in company via the less familiar St Peters route. The Squirrel, being the faster of the two aircraft, flew in the lead.

Nearing the end of the flight, both aircraft were asked to hold for ten minutes at Rose Bay. Both pilots later said that they became concerned because it was almost at the end of daylight, and neither helicopter was certificated to undertake commercial night operations. After holding for about seven minutes the tower cleared both aircraft to track to Sydney Airport via St Peters and requested the pilots to expedite their return.

Whilst tracking inbound, the Jetranger began to fall behind the lead helicopter. The pilot of the Jetranger did not advise the tower, nor the pilot of the Squirrel, that he had fallen behind. As he approached to cross runway 25, the pilot of the Jetranger observed a fixed wing aircraft on final for that runway and slowed his helicopter because of wake turbulence concerns. The tower, on hearing the Squirrel report on the ground at Helipad H1 on the southern side of runway 25, assumed that the Jetranger was also about to land. At the same time a Sikorsky S-76A helicopter was inbound from the Harbour Bridge for Helipad H4, on the northern side of runway 25. A short time later the pilot of the Sikorsky reported that a Jetranger had passed in front of him from right to left, within a distance of about 150 ft, near Helipad H4.

All three pilots subsequently reported on the difficulties of sighting other aircraft against the city lights. Although the pilot of the Squirrel was aware of the proximity of the Sikorsky, the pilot of the Jetranger was not. Nor was the Sikorsky pilot aware of the proximity of the other two helicopters. The aerodrome controller did not see the Jetranger against the background of aircraft and city lights and was consequently unaware of the potential traffic confliction.

The pilot reported that he started the pre-landing checks as he joined the circuit and not on the downwind leg which is where he would normally start them. He selected full flap earlier than normal on the base leg and checked the throttle was at idle because the aircraft was not descending as quickly as he expected. He could not work out why. At no stage did the landing gear warning horn sound. After he flared the aircraft for the landing the propeller began to strike the runway, and the aircraft settled onto its fuselage. It slid to a stop to one side of the runway. The pilot indicated that he had forgotten to select the gear down prior to landing and had not noticed the lack of gear indications during his checks.

The pilot further advised that the gear warning horn had been reported as unserviceable on previous flights and although he had been told that it had been repaired it was apparently still unserviceable.

One factor often identified during inadvertent wheels-up accidents is the interruption of the pre-landing checks, usually because they have been started too early and the pilot is not ready to extend the gear when he gets to that item on the list. Universally recommended practice is to restart the checks from the beginning if they have been interrupted for any reason. The pilot in this accident reported that he had started his checks early and had stopped at the gear item.