Helicopter VH-AUQ was substantially damaged in a ground resonance occurrence in November 1995. It was rebuilt and released, as serviceable, on 13 April 1996. The helicopter was immediately flown from Australia's east coast to Fitzroy Crossing, arriving there on 24 April 1996. During the flight to Fitzroy Crossing, which took 24 flying hours, the pilot/owner reported that, on one occasion, there had been some vibration in the rotor system. The vibration occurred after touchdown and immediately prior to shutdown at Mt Isa, it was minor and had only lasted for a short period.

The helicopter was fitted with a single set of controls. On the day of the accident another pilot attempted to fly the helicopter, whilst the pilot/owner sat in the passenger's seat. After start up, and prior to lift off, it began to vibrate. The helicopter was shut down and inspected. No reason for the vibration was found. After some discussion it was decided that the vibration had been caused by incorrect positioning of the cyclic control stick during start up. The pilot/owner re-phased the main rotor blades, i.e.. positioned the blades evenly around the rotor head, before attempting another start. The pilots then switched seats and strapped in. The pilot/owner started the engine and completed a power check. He reported that there was no vibration and everything appeared normal.

The pilot flew the helicopter to a 1 m hover and commenced a turn to the left. As the helicopter started to yaw a severe vibration started. The vibration was so bad the pilot was unable to maintain a reasonable degree of control and he reduced power and allowed the helicopter to touch down. The vibration continued, increasing in amplitude. This made it difficult for the pilot to reach the mixture control lever. However, he was finally able to shut the engine down. During the vibration after touchdown, the bubble fractured across the top and bottom frames and fell forward. The pilot/passenger's seat belt came undone, and she was ejected forward of the helicopter, falling onto a grass area. By the time all movement had stopped, in addition to the bubble damage, the tail boom had been cut off, most of the airframe structure had been fractured or severely distorted and the main and tail rotor blades destroyed. The pilot/owner was assisted from the wreckage by an observer.

The investigation found that lockwire and spacing washers were missing from the torque-nuts on all three main rotor blade dampers. Each of the torque-nuts was found in a different position on its shaft. None of the nuts were positioned near where they were expected to be on a newly overhauled blade damper. Evidence indicated that all three dampers were overhauled prior to being fitted during the rebuild. Two of the blade dampers were severely damaged and their torque setting could not be checked. The third damper was undamaged, and its torque was found to be 140-inch pounds counterclockwise and 135-inch pounds clockwise. The manufacturer's minimum torque requirement is 230-inch pounds, in both directions on installation and 200-inch pounds in service. The internal bearings and friction plates, in all three dampers, were in good condition. All three damper arm uni-ball bearings, which had been replaced during the overhaul, were found to have excessive play. Two of three bolts, which were used to secure linkages between the trailing edge of the main rotor blades and their respective dampers, were found to be badly worn, as were their corresponding lower bolt holes. Whilst the play in the uni-ball bearings was probably the result of wear during the accident-flight vibration, the extent and condition of the wear on the bolts indicates that it had been there for some time.

The lack of torque and the wear on the bolts and holes probably led to partial and uneven damping of lead and lag in the main rotor blades. This, in turn, led to the vibration which caused the accident.

The organisation which overhauled the blade dampers reported that it had been completed in accordance with the manufacturer's requirements. The dampers had been delivered to the rebuild-organisation with spacers and lockwire fitted and torqued to the correct setting. The rebuild-organisation reported that they had fitted the dampers, as delivered, and had not had to make any adjustments to them prior to releasing the helicopter to the owner. The pilot/owner reported that the lockwire had been removed during adjustment and ground testing, after the dampers had been fitted to the helicopter and before it was released to him. He had assisted whilst adjustments were made to the dampers by maintenance personnel. The spacers and lockwire were not refitted.

It could not be determined how worn bolts came to be fitted to the damper linkages. Neither the overhaul- or the rebuild-organisation believed it had fitted the bolts.

It is possible that the dampers were correctly torqued when the helicopter was released to the owner. Although no evidence of a similar occurrence is available, it is possible that the absence of lockwire allowed the torque-nuts to back off gradually, whilst the helicopter was being flown to Fitzroy Crossing. The reducing damper torque went unnoticed by the pilot/owner, probably because of his inexperience. The different positions of the torque nuts, at the time of the accident, indicates that, if the nuts had backed off, they had done so unevenly. The alternative explanation is that the blade damper torque-nuts were incorrectly set during adjustments made before the helicopter was released to the pilot/owner. However, had this been the case any significant variation in damper torque would have been evident as vibration in the rotor system, much earlier than reported. Therefore, the first alternative is considered the more probable explanation and the uneven and reducing damper torque probably reached a point, on the last two flights, where the vibration became excessive.

An inspection of the pilot/passenger's seat belt tongue and buckle indicated there was a significant amount of movement between the tongue and the buckle when the tongue had been pushed into the locked position. This made it possible, at some angles, for the tongue to be pulled out again, even though the belt appeared to be locked.

During a cattle mustering operation, the helicopter suffered a total power loss at about 80 ft AGL. The resultant heavy landing spread the skids, and the main rotor severed the tail boom. Both occupants were able to exit the aircraft safely.

Post flight examination found there was no fuel in the helicopter's tanks. The tanks were intact.

The pilot has not responded to a request for information.

On arrival overhead, the pilot observed sheep on the flight strip. He made two low passes in an attempt to clear the sheep. During the third pass the remaining sheep cleared the runway so the pilot decided to discontinue the pass and make a straight in approach for a landing.

The pilot did not complete his pre-landing checks, and the landing gear was not lowered. The landing gear warning horn did not sound during the approach, possibly because of a higher than normal power setting.

The aircraft was landed with the landing gear retracted.

The pilot reported that he was distracted by the need to chase the sheep of the strip, and he forgot to lower the gear.

The Mitsubishi MU-2 was to be flown on a cargo flight from Bankstown to Wagga Wagga. During his pre-flight inspection the pilot cleaned the windshields with warm water to defog them. The Swearingen SA226 was also to be flown on a cargo flight from Bankstown to Cootamundra. The pilot of that aircraft wiped the side windows with a chamois, as the front windshields were already clear.

The MU-2 taxied at 0630 for runway 29C, followed some three minutes later by the SA226. The MU-2 proceeded along taxiway X, then across the 29 engine runup area towards taxiway S1, which also required crossing taxiway K. The SA226 proceeded along taxiway K, towards taxiway S1. As he taxied through the runup area the MU-2 pilot was unaware of the presence of the SA226 on taxiway K. The SA226 pilot had seen the MU-2 to his left, taxiing through the runup area, and assumed that the pilot of that aircraft would give way to him. He reported that his attention was mainly directed ahead and to his right, as he monitored departing aircraft.

As the MU-2 was turning left out of the runup area to cross taxiway K, the pilot felt a lurch to the left as his aircraft was struck from the rear right. The wing of the SA226 had passed under the right-wing tip fuel tank and then into the propeller of the right engine, resulting in substantial damage to both aircraft. The pilot of the SA226 had been unaware of the close proximity of the MU-2 until the impact pulled the nose of his aircraft to the left, and he saw the other aircraft. Both pilots conducted emergency shut-downs and escaped from their aircraft without injury.

Pilots of adjacent aircraft witnessed the accident and reported that it was a clear morning with no obstructions to visibility. Immediately prior to impact, the witnesses observed the aircraft to be converging at an angle of approximately 50 degrees and thought it was going to be close. When the impact was seen as inevitable, they were unable to give a warning call as the ground radio frequency was congested.

The two aircraft were on crossing tracks. VH-NJG was cleared to descend initially to flight level (FL)180 to maintain separation with the VH-EKH at FL170. The controller noted successive radar returns indicating NJG was at FL177 and assumed the pilot had continued descent through the cleared level. He decided that the most appropriate course of action was for the aircraft to continue descent through the level of EKH while maintaining separation laterally.

He therefore cleared NJG to descend to FL115 and instructed the aircraft to alter heading 15 deg to the left. However, the groundspeed and descent rate of NJG did not proceed as anticipated and it became evident to the controller that vertical separation would not be established before lateral separation was lost. Accordingly, the controller instructed NJG to expedite descent through FL160 but did not pass traffic information to either aircraft when separation was lost. Radar separation reduced to three miles, instead of the required five miles, before vertical separation was established.

The investigation revealed that when the controller noted NJG's radar return indicating FL177, he assumed the pilot had continued descent through the cleared level. He then decided the conflict would be most quickly resolved by the aircraft continuing descent to below FL160, rather than reversing its flight profile to climb back to FL180. This was why he cleared the aircraft to continue descent to FL115. A replay of the radar tape indicated that the aircraft actually levelled out at FL180 but in doing so had gone down to FL177 in the level out manoeuvre (there were three returns indicating FL177 before FL180 was indicated). Hence the descent instruction resulted in the exact effect the controller was trying to avoid which was a delay due to reversal of flight profile while vertical separation did not exist.

FACTUAL INFORMATION

A Boeing 747 from Sydney, was on descent for landing at Cairns airport. The duty runway was runway 15 and the weather was visual meteorological conditions (VMC). The aircraft was inbound from the south-east and the controller elected to track the Boeing 747 in close proximity to the departure end of the active runway while manoeuvring for a landing from the north. The crew of the Boeing 747 had been assigned progressively lower levels south of the airport and had reported visual. Subsequently, the approach controller instructed the crew to descend to "3,000 ft visual" while the aircraft continued to track direct to the Cairns very high frequency omni-direction radio range (VOR) navigation aid.

Due to pending departures from runway 15, the approach controller planned to maintain the Boeing 747 at 3,000 ft until the mid-downwind position. This would have enabled the departing aircraft to climb to an initial level of 2,000 ft, while ensuring vertical separation from the inbound Boeing 747. Once radar or visual separation was established between the two aircraft, the approach controller intended to instruct the respective crews to climb or descend as appropriate. The controller had not specifically advised the Boeing 747 crew of any expected departures but had on three occasions while the aircraft was south of the airport, instructed them to "Maintain 3,000 ft visual".

When the Boeing 747 was immediately south of the airport, the approach controller requested the crew to report sighting a De Havilland Twin Otter aircraft outbound to Lizard Island (to the northeast) on climb to 3,000 ft. The controller was concerned that radar separation between the Boeing 747 and the Twin Otter may have been reduced as the Boeing 747 tracked downwind. Assigning visual separation responsibility to the crew of the Boeing 747 would ensure separation was maintained. The crew of the Boeing 747 reported sighting the Twin Otter and advised that they could pass behind the aircraft while tracking downwind.

At this stage, the crew were under a high workload due to the higher-than-normal altitude, the proximity to the aerodrome, and the requirement to configure the aircraft for landing. The controller instructed the crew of the Boeing 747 to track downwind, but did not re-confirm the requirement to maintain 3,000 ft or assign a visual approach; nor was the controller required to do so. At the same time, the pilot of a Britten Norman Islander aircraft departing from runway 15 had been instructed to maintain 2,000 ft to ensure vertical separation with the Boeing 747. As the Islander aircraft became airborne, the approach controller noticed that the radar display altitude label for the Boeing 747 was approaching 2,500 ft (on descent), as the aircraft entered early left downwind. After the controller queried the aircraft's level with the crew and confirmed that they were on descent, he instructed the crew to maintain 2,000 ft. However, as the Islander aircraft had not climbed above 1,000 ft and visual separation was being used by the controller, the crew of the Boeing 747 were then instructed to make a visual approach. The Boeing 747 landed shortly after without further incident. There was a breakdown in procedures but no breakdown in separation.

ANALYSIS

The Boeing 747 crew were not aware of the pending departure of the Islander aircraft and, on being instructed to track for downwind, their perception was that they had been issued with a visual approach instruction.

Consequently, the crew commenced descent to circuit altitude. Pilots are required to maintain the last assigned altitude or flight level until cleared to a lower (or higher) altitude, cleared to carry out a visual approach, or cleared for final when conducting an instrument approach. In this case no such instruction had been issued. The instruction to "Maintain 3,000 visual" only indicated that responsibility for terrain clearance had been assigned to the crew.

The crew developed an expectation of a clearance for a visual approach after sighting and being required to visually pass behind the outbound Twin Otter. Additionally, the crew were concerned with the need to slow the aircraft to circuit speed while having to descend to circuit altitude and complete cockpit checks in time for landing. The crew's haste to complete a number of tasks concurrently pre-disposed them to making an error. The crew could have notified the approach controller of its concerns and requested radar vectors or alternative tracking to ensure sufficient time to complete tasks prior to landing. However, they became focused on the tasks within the cockpit, and sighting and avoiding the outbound Twin Otter.

The crew may have been alerted to the situation had the controller provided more information. If the controller had notified the crew to expect descent below 3,000 ft on downwind due to departing traffic from runway 15, they may have concentrated on maintaining the assigned level until specifically cleared to make a visual approach. While the approach controller did provide separation and was not obliged to provide traffic information to the crew of the Boeing 747, the lack of traffic awareness was a failed defence in the air traffic system. Alternatively, the controller could have reiterated the requirement to "Maintain 3,000 ft visual" with the new tracking clearance for downwind. Either of these actions may have provided a defence against the incident.

A compounding factor in the incident was the current Aeronautical Information Publication (AIP) radiotelephony phraseology for flights in controlled airspace. The phrases in AIP OPS CTL - 38 paragraph 31 are ambiguous and open to misinterpretation. Specifically, the phraseology relates to assignment of separation responsibility and could lead a pilot to expect a clearance to make a visual approach, after traffic reported by a controller is sighted by the pilot. This is not necessarily the case, as often the sighting of traffic by pilots, is only one aspect of separation in the airspace management process. Depending on the situation a number of methods of separation can be used to ensure the separation of one aircraft from a number of other aircraft.

SIGNIFICANT FACTOR

The crew of the Boeing 747 misinterpreted the controller's instruction to track for left downwind as a clearance to make a visual approach and, consequently, descended below the last assigned level of 3,000 ft.

SAFETY ACTION

Local Safety Action

1. The operator issued an operations flight standing order which cautioned crews to be vigilant to the differences between a clearance to "Descend to .... thousand visual" and "Make visual approach".
2. The operator has proposed publishing an article relating the aspects of the incident in the company safety magazine.

Bureau of Air Safety Investigation Action

As a result of this investigation the Bureau of Air Safety Investigation issued the following interim recommendation to Airservices Australia on 7 November 1996:

"IR960093

The Bureau of Air Safety Investigation recommends that Airservices Australia review the wording in AIP/OPS CTL-38 and MATS 12-A-18 to prevent possible confusion regarding a clearance to make a visual approach".

The aircraft was on a private flight to Katoomba. As the airstrip was not licenced, the pilot rang the operator for permission to land there. He was not advised of any hazards at the airstrip.

On arrival in the circuit area, the pilot determined that the wind conditions favoured landing in the 06 direction and observed power lines crossing the approach path. The pilot reported that he decided to land the aircraft further along the runway than normal, to ensure adequate clearance from the power lines during the approach. After landing, he applied the brakes and steered the aircraft to the left side of the strip, prior to making a right turn to backtrack. However, the nose wheel struck a deep rut near the side of the strip, collapsing the nose landing gear, and allowing the propeller to strike the ground.

FACTUAL INFORMATION

A Fokker F28 aircraft was conducting pilot training in the Perth terminal area. The terminal information indicated that runway 03 was to be used for departure tracks to the west of the extended runway 21/03 centreline and runway 06 for other departures. Runway 03 was to be used for arriving aircraft. The weather was CAVOK (No cloud below 5,000 ft and with a visibility greater than 10 km). To assist with the management of the training aircraft, the aerodrome controller (ADC) had received an airspace release within 5 NM of the aerodrome to the east of the centreline of runway 21/03 and up to an altitude of 1,500 ft. Traffic levels were moderate and the runway configuration increased the complexity of the traffic sequence. The control tower was normally manned by three air traffic controllers. One of the controllers was required to leave the tower and there was no replacement available. The control tower was then manned by the ADC and the tower co-ordination (COORD) controller.

The F28 had completed an instrument landing system (ILS) approach to runway 24. The crew of the F28 requested to overshoot to the left from the approach and then to conduct a right circuit to runway 03. The crew were cleared to overshoot to the left on climb to 1,500 ft and to remain east of the centreline of runway 21/03. The crew of the F28 complied with the clearance and remained east of runway 21/03.

An international Boeing B747 aircraft was being radar vectored for positioning on a left base for runway 03 and a Cessna Conquest C441 was being radar vectored from the east of Perth for a right base to runway 03. The pilot of the C441 had been cleared to descend to 2,500 ft. The ADC determined that there would be insufficient time for the F28 to land on runway 03 before the B747.

The ADC decided to hold the F28 to the southeast until the B747 had landed. The ADC's intention was to then instruct the crew of the F28 to continue for a landing on runway 03. The approach radar east (APPE) controller required the F28 to leave the circuit area to enable the C441 to descend for landing and to assist with the approaches of three following aircraft. The APPE controller co-ordinated with the ADC for the F28 to track to Parkerville, located to the northeast of the aerodrome, on climb to 2,500 ft for another ILS and for the C441 to be positioned on a close right base for runway 03. The ADC agreed with the proposal.

The APPE controller was aware that the F28 and the C441 may conflict. The APPE controller used non-standard phraseology to instruct the ADC to not approve the climb to 2,500 ft for the F28 until the C441 was closer to the aerodrome. The APPE controller also requested advice from the ADC of when the C441 could descend below 2,500 ft. The two controllers did not establish who was to be responsible for separating the F28 and the C441, or co-ordinate suitable clearances which would provide separation assurance between the aircraft.

The ADC believed that if the crew of the F28 received the instruction to track to Parkerville immediately, that the separation with the C441 would be maintained. The ADC instructed the crew of the F28 to track to Parkerville and to climb to 2,500 ft.

After observing the landing of the B747, the ADC returned his attention to the F28. The ADC did not request the COORD controller, who was a rated ADC, to assist in the monitoring of aircraft. The ADC noticed that the F28 had tracked further to the south than expected and that this had placed the aircraft in conflict with the C441. The ADC attempted to contact the crew of the F28 but was unsuccessful, as they had previously been instructed to call the APPE controller. Visual separation from the tower could not be used due to the proximity of the aircraft to each other. The APPE controller instructed the pilot of the C441 to climb to 3,000 ft and to turn to the south for separation. The aircraft passed with approximately 1 NM horizontal separation and 500 ft vertical separation. The required separation was 3 NM horizontally or 1,000 ft vertically. There was a breakdown in separation.

ANALYSIS

As the aircraft would possibly conflict at a point close to the boundary of both controller's area of responsibility, the ADC and the APPE controllers needed to co-ordinate a separation procedure to ensure separation was maintained between the inbound C441 and the outbound F28. The controllers also needed to establish which of them would undertake responsibility for the separation and for both aircraft to be transferred to that controller's radio frequency for the application of separation.

The co-ordination between the controllers did not adequately address these aspects and consequently there was a lack of separation assurance.

The ADC could have utilised the COORD to monitor some of the traffic during the period when visual separation was to be applied. The ADC possibly became distracted with other traffic and did not adequately monitor the flight of the F28 with reference to the C441.

SIGNIFICANT FACTORS

1. The ADC and APPE controllers did not co-ordinate an adequate separation procedure before transferring each aircraft to the other control position.

2. The ADC did not utilise the COORD controller to assist in the monitoring of aircraft.

The Boeing 737 was northbound from Melbourne to Brisbane, via route H29. The air traffic controller cleared the Boeing 737 direct to Mudgee so that it would pass about 15 nautical miles west of Wagga. The Boeing 747 was eastbound from Perth to Sydney on route H31 tracking from Natya to Cullerin. This track crosses H29 about 20 miles north of Wagga. Both aircraft were maintaining flight level 370 as cleared by air traffic control.

Less than three minutes before the Boeing 747 reached the crossing point of the two tracks, the air traffic controller instructed the aircraft to descend to flight level 350. At the time the Boeing 747 commenced the descent, recorded radar data showed that the aircraft had a groundspeed of 530 knots and was 24 miles from the point of track intersection. At the same time the Boeing 737 had a groundspeed of 480 knots and was 20 miles from the point of track intersection. Subsequently both aircraft passed with 1,600 ft vertical separation and within 3 miles horizontally.

The time allowed by the controller for the descent was inadequate to ensure the required 2,000 feet vertical separation standard for aircraft within five miles of each other was maintained. The controller's resolution of the conflict was deemed to be inappropriate for the situation. Given the short time to the crossing point, a higher degree of expedition would have been required if descending the Boeing 747 was to be the only measure to achieve separation. For separation assurance, once the conflict was recognised, it was not enough for the controller to merely instruct the Boeing 747 to descend. A requirement or a radar vector was necessary given the proximity of both aircraft.

Weather was not a factor in the incident. The air traffic controller's workload was not busy. All facilities and equipment were serviceable. The controller was licensed and current. There was no evidence that the controller was suffering fatigue.

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

1. The air traffic controller did not adequately plan ahead to ensure required separation standards were achieved.

The pilot reported that he made an unplanned diversion to Moorabbin due to unsuitable weather en route to his destination. His approach to runway 17 right was "on the low side" and late on the approach the airspeed decreased, and the sink rate increased. He applied power but this did not arrest the sink rate. The aircraft touched down heavily in soft earth short of the commencement of the sealed runway. This caused the nose landing gear assembly to collapse and then the nose section forward of the first bulkhead broke off as the aircraft slid onto the sealed runway. The aircraft was stopped in the grass on the right side of the runway.

The wind velocity reported by the tower at the time of the accident was from 200 deg at 15 to 20 kt.