The pilot of the Hiller 12E helicopter had been tasked with spraying insecticide on a paddock of soybean plants. The pilot familiarised himself with the property layout and obstacles before starting treatment of the area. The paddock to be sprayed was crossed at its mid-point by powerlines running east-west. The southern boundary of the paddock was marked by a wire fence, a road and another set of powerlines.

The pilot reported commencing a spray run along the northern boundary but found that his GPS navigation system was not working. He landed on the western side of the paddock to re-enter data into the GPS system. On carrying out another run, the pilot found the system was still not operating, so he moved the helicopter away from the paddock to the south. He then elected to return to the paddock by flying over the powerlines. The pilot reported that at this time, he was mentally occupied with the problem with the GPS and forgot about the powerlines at the southern edge of the paddock. The helicopter spray gear contacted the powerlines, causing the helicopter to pitch nose down and contact the ground. The helicopter came to rest inverted, and the pilot climbed out of the wreckage, receiving only minor injuries. The helicopter did not catch fire.

A significant factor in this accident was the preoccupation of the pilot with the GPS navigation equipment at a time when he was flying close to obstacles.

The pilot was preparing to depart on a private flight from Cobden Vic. to Moorabbin. The Piper Cherokee aircraft was parked in front of the club house when the pilot found that the starter motor would not engage. After various attempts to get the starter motor to engage were unsuccessful the pilot decided to hand swing the propeller.

The pilot advised that he applied the parking brake, set the throttle to approximately half power and turned the magneto switches on. When he swung the propeller the engine started and ran at high RPM. The parking brakes did not stop the aircraft from rolling forward. The pilot grabbed the wing to try to arrest the aircraft's progress, however it continued to gather speed. When the pilot could no longer hang onto the wing he grabbed the tail plane. The pilot was carried 300 metres before he relinquished his grip and the aircraft sped away.

The aircraft continued south past the windsock and broke through the perimeter fence into a paddock. It ran across the paddock and through another fence and across the airfield access road. On the far side of the road, it demolished a gate before colliding with a shed and toilet, coming to rest against a farm plough. The aircraft had travelled in excess of 1 1/2 km.

The low time pilot had not been trained in hand starting aircraft engines. He said that he was not aware of the regulatory requirements surrounding the starting of aircraft engines, and did not ask a nearby pilot to sit on the brakes while he attempted the hand start. He stated was not aware that the operator from whom he rented the aircraft specifically banned unauthorised attempts to rectify defects when away from Moorabbin, including the hand starting of aircraft engines.

FACTUAL INFORMATION

Following a request from an Australian Army (Army) unit for the use of the Port Wakefield proof firing range (designated as restricted area R295 A to F), the Defence Corporate Support Centre (Support Centre) in Adelaide mailed notices to a number of organisations on 3 April 1998. These local procedures included notification to the Royal Australian Air Force (RAAF) at Edinburgh for the issue of a notice to airmen (NOTAM). R295 A, B and E were required for missile firings on 6 and 7 April 1998 and would be active from the surface to 33,000 ft. The lowest useable level by overflying aircraft was 34,000 ft. The letter to RAAF Base Edinburgh was not received prior to the period nominated for the activity and a NOTAM was not issued.

On Monday 6 April 1998 the Army unit arrived at the proof firing range and was cleared by the range control officer (RCO) to conduct their exercise from 1200 Central Standard Time. When the requesting unit arrived at the range, the RCO assumed that a NOTAM had been issued and that the other agencies had been notified.

At approximately 1230 and 1445, Boeing 737 aircraft operating scheduled passenger flights from Sydney to Perth on air route Q32 passed through the northern portion of R295E at flight level (FL) 280 and FL310 respectively. During the period between the range opening and detection of the second B737 overflying the range, the unit had fired two target missiles. The B737s had flown through an active firing range.

Detection of the B737

The RCO heard the second B737 as it flew through the area and cancelled the exercise for the Army unit. At 1455, he rang Adelaide Air Traffic Control (ATC) to confirm that the range was active to 33,000 ft. Subsequently, ATC ascertained, and advised the RCO, that a NOTAM had not been issued for the activity.

Notification procedures

Army procedures required the unit intending to operate on the range to notify the Support Centre. The Support Centre was to advise the appropriate airspace authority prior to the activity for the issue of a NOTAM. The Support Centre used RAAF Edinburgh as the appropriate authority for the issue of a NOTAM for a range activity. The orders did not specify a minimum period of notice to the Support Centre. The Manual of Air Traffic Services (MATS) required a restricted area NOTAM to be issued a minimum of 8 hours prior to the commencement of operations.

There were no procedures requiring the unit intending to use the range to confirm that the activity had been notified in accordance with Army procedures. The RCO was not required to check that a NOTAM and notices to other authorities had actually been issued prior to approving operations on the range.

The procedures did not require the RCO or the unit to confirm the commencement or completion of operations with any air traffic service (ATS) agency.

Range surveillance

The RCO had radar to assist in the surveillance of that portion of the range over the Spencer Gulf. The radar was capable of detecting vessels on the water and persons on the tidal flats but was not designed to detect aircraft at high altitude. Because of these design and operating mode limitations, the radar did not detect the B737s. The range staff did not maintain a record of unauthorised entries to the range.

ANALYSIS

The process for requesting activation of the range was reliant upon staff members completing actions, which subsequently transferred the responsibility for completion to another person. Failure to complete an action by any one position in the required sequence of events would result in the notification process, for range activation, to stop. This process had no checks to ensure that subsequent actions had been completed or for feedback to occur if any action did not happen. Overall, the procedure was a fail-unsafe process. A final check with the relevant ATS agency prior to clearing the unit to operate on the range would have alerted the RCO to the fact that a NOTAM had not been issued.

The RCO was reliant on visual and limited electronic surveillance of the range area which was not conducive to the detection of high-flying aircraft. A check with the ATS agency would have established if any aircraft were operating within the range at high altitude.

SIGNIFICANT FACTORS

1. Range procedures did not require the Support Centre or the unit requesting the use of the range to confirm that a NOTAM had been issued.

2. The Support Centre did not allow sufficient time for the notice to be received and for a NOTAM to be issued.

3. Range procedures did not require the RCO to confirm, with the relevant ATS agency, that a NOTAM had been issued or whether there were any aircraft in the area prior to approving operations on the range.

4. Range surveillance facilities were unlikely to detect aircraft operating at high altitude within the range.

SAFETY ACTION

Local safety action

The Support Centre has amended procedures to require 21 days notice of range activities and will notify the appropriate authority 14 days prior to the activity for submission of a NOTAM. Additionally, the appropriate authority will fax a copy of the request for a NOTAM to the Support Centre as confirmation. Furthermore, the unit intending to operate on a range must access the aeronautical information service and obtain a copy of the current NOTAM for the activity prior to commencing operations. RCO clearance for operations on a range is subject to the unit providing a copy of the current NOTAM.

Bureau of Air Safety Investigation safety action

As a result of the investigation of this and another similar occurrence, the Bureau of Air Safety Investigation issued the following interim recommendations to the Australian Defence Force and to Airservices Australia on 30 June 1998:

"IR980086

The Bureau of Air Safety Investigation recommends that the Australian Defence Force, with the assistance of Airservices Australia, review airspace activation procedures to ensure that appropriate and fail-safe notification actions have been completed prior to the commencement of military activities.

IR980087

The Bureau of Air Safety Investigation recommends that Airservices Australia assist the Australian Defence Force to review airspace activation procedures to ensure that appropriate and fail-safe notification actions have been completed prior to the commencement of military activities".

The following response to the interim recommendation was received from Airservices Australia on 27 July 1998:

"I refer to your letter of 29 June enclosing Interim Recommendation IR980087. The subject of notification and activation of areas for Military activity will be referred to the National Airspace Management and Air Traffic Services Sub Committee (NAMPS) of the Air Coordinating Committee (ACC) for resolution. As a joint Australian Defence Force/Airservices Australia forum, NAMPS is the appropriate body to review airspace activation procedures relating to military activity".

Response classification: CLOSED - ACCEPTED

The commercial pilot pre-arranged to private hire a Robinson R22 for business purposes. When he arrived at Mangalore to collect the helicopter he spoke to the duty instructor who advised him to take VH-KKG after it returned from a training flight. Minutes later a solo student pilot landed KKG well away from the AVGAS refuelling facility because the flying school forbade student pilots to hover or land near the fuel bowsers. The student then applied the control frictions and reduced power, expecting to stop the engine after its cool down period. The commercial pilot planned to depart Mangalore with full fuel.

He proceeded to the right door of KKG where he tried to explain to the student that he wanted to swap seats with him without the engine being shut down, in order to hover-taxi to the bowser for fuel. The student, a Japanese national undergoing basic flying training in Australia, did not understand the request mostly because of the over-riding engine noise and because he was wearing a headset. He remained at the controls with the engine running. As the commercial pilot intended to reposition the helicopter only a short distance before shutting down the engine to refuel, he proceeded to the left side of the helicopter and occupied the left seat where there were dual flight controls.

He explained his intentions to the student who handed over the controls. The commercial pilot then checked that both his and the student's seat belts were fastened before releasing the control frictions. He then increased the engine/rotor RPM to the top of the green and raised the collective lever with the intention of stabilising in the hover momentarily before hover-taxiing to the bowser. He reported that as the helicopter became light on the skids, he applied light pressure to the anti-torque pedals and made minor cyclic control inputs as he further raised the collective.

As the landing gear skids left the ground the helicopter yawed quickly to the left. He cannot recall exactly what his following control inputs were as things happened very quickly and violently. However, his impressions were of yawing quickly and an unusual attitude which he attempted to correct. Then the helicopter skids contacted the ground and it lifted back into the air momentarily before settling onto the ground and rolling onto its left side. The weather was reported to be fine with a light and variable wind, CAVOK and temperature about 25 deg Celsius. The weather was not a factor in the accident.

No fault has been subsequently found with the helicopter which may have contributed to the accident. It had been functioning normally on its previous flight. The student reported that he made no flight control input after handing over the controls to the private pilot. The commercial pilot recalled that the student grabbed the cyclic at the end of the accident sequence but this action did not influence the outcome of the accident. The instructor did not give the commercial pilot approval for a engine-running seat swap with the student, nor did he authorise him to fly from the left seat.

Prior to the accident, the instructor had been very busy escorting a group of cadets who had arrived at the airport earlier than expected to visit the flying school. Except for dual check and training flights, the helicopter manufacturer requires that the R22 be flown from the right seat. There is a different visual perspective flying from the left seat versus the right seat. The R22 is also equipped with the RPM governor on the right throttle only. Normally only check and training pilots are given the additional dual instruction to enable them to fly from the left seat with another pilot occupying the right seat.

Commonly, when a pilot first attempts to hover a helicopter from the opposite seat, some degree of over-controlling will occur. The commercial pilot reported that he had previously controlled an R22 in flight from the left seat with a pilot-in-command occupying the right seat, but he had never before lifted a helicopter into the hover while flying from the non command seat. It is probable that the commercial pilot began to over-control the helicopter as it lifted into the hover. He was experienced at flying without a governor controlling RPM, so manipulating the left throttle was probably not a factor in the accident.

The yaw to the left probably resulted from inadvertent excessive left pedal input because a loss of anti-torque or insufficient left pedal would have caused the helicopter to yaw to the right. Over-controlling probably developed into an irretrievable roll-over condition as the left skid dragged on the ground as the helicopter yawed left.

The Boeing 737 (B737) was maintaining flight level (FL) 350 and the Melbourne Central controller had completed all the required tasks involving that aircraft. Although the aircraft had not yet reached the point at which it would normally be handed off to the next sector (Sector 4), the controller decided to proceed with an early hand-off which was accepted by the Sector 4 controller. The Airbus A320 was maintaining FL370 on a track which crossed that of the B737. The appropriate vertical separation standard of 2,000 ft was being maintained. As the tracks of the aircraft converged, the crew of the A320 was given descent to FL350 by the Melbourne Central controller, who momentarily forgot about the B737.

As soon as the crew of the A320 reported leaving FL370, the controller realised that he could not guarantee the required horizontal separation standard of 5 NM and immediately instructed them to return to FL370. The A320 had descended to FL368 before the crew were able to arrest the descent and commence climb back to FL370 and re-establish a vertical separation standard. The investigation revealed that the aircraft were at the same flight level, with a horizontal separation of 4.7 NM, resulting in a breakdown of separation standards. The practice of early hand-off is being reviewed by Airservices Australia as it can lead to a reduced level of awareness by controllers.

The pilot in command was demonstrating a 15-degree banked turn at a height of about 700 ft. The aircraft was travelling at 70 kts with 15 degrees of flap extended. However, control of the aircraft was affected by an unseen dust devil which caused it to roll further left to about 90 degrees angle of bank. The pilot applied full right rudder and aileron controls, righting the aircraft just above treetops in a stalled attitude, but was unable to arrest the descent. The aircraft struck trees and came to rest on the ground in an inverted position. Both the pilot and passenger vacated the aircraft, which was subsequently destroyed by an ensuing fire.

The pilot of the aircraft reported that the helicopter was being used to spray a sugarcane crop when he attempted to avoid a bird. As he banked the helicopter to the right, he heard two bangs and decided to land the helicopter in an adjacent fallow paddock. During the attempted landing, the helicopter yawed right, and the right skid dug into the ground, rolling the helicopter onto its right side. The pilot was not injured. Examination revealed that the right spray boom had struck the crop during the attempt to avoid the bird.

A portable ELT was carried in the cabin, but it was not activated by the pilot.

The pilot of a motorised glider declared an inflight emergency to Flight Service, advising that he was above cloud, south-west of Camden, with 50 minutes fuel endurance. The aircraft was not equipped with a transponder but carried a Global Positioning System. An alert phase was declared and, following liaison with Sydney Approach, the aircraft was located by primary radar returns. The alert phase was later upgraded to a distress phase when it was apparent that the aircraft would shortly run out of fuel while still above thick cloud.

Communications with the aircraft ceased until a 'Careflight' helicopter finally made contact with the pilot, who indicated that the aircraft had run out of fuel and was gliding. The 'Careflight' pilot subsequently reported that the glider had landed, and the pilot was uninjured. The glider pilot later reported that he had planned to fly from Gulgong to Camden. The weather forecast indicated that the weather on the coast was poor, with extensive low cloud. However, by his own observation, the weather was fine to the south of Gulgong. The aircraft was winch launched to a height of 1,500 ft at Gulgong before departing to the south with an almost full fuel tank (22 litres maximum), with the pilot intending to divert to Bathurst if necessary.

The weather south of Bathurst continued to improve, but low cloud cover extended to the east towards the coast. By the time he had thermalled to 9,500 ft the glider pilot said he was about 80 km west of Camden and some 10 km west of the line of cloud. After contacting Camden tower for the current weather situation, the pilot assessed that he could continue in accordance with the Visual Flight Rules and land at Camden. After gliding down to 6,500 ft, the motor was started and he continued towards Camden, above cloud. However, further contact with Camden tower indicated that the weather had deteriorated and there was now almost total cloud cover.

The pilot contacted Sydney Centre for a clearance into controlled airspace and held for some time about 20 km west of Camden to assess the weather. As there was no improvement, he was authorised to climb to 10,000 ft so that he could then glide to the west, beyond the cloud line, into suitable outlanding country. At 10,000 ft the engine stopped when the fuel was exhausted. However, the engine could not be retracted by normal or emergency methods, severely restricting the glide performance of the aircraft. Unable to now reach the end of the cloud line, and descending into wispy cloud, the pilot finally found a hole in the cloud and spiralled down to land in a small paddock.

After touchdown the pilot had to ground loop the glider to avoid colliding with a fence, resulting in damage to the landing gear. The pilot subsequently reported that after checking and re-crimping the electrical wiring for the engine retract motor, the engine retracted normally.

While landing towards the north-west at Bruny Island, the aircraft landed heavily and bounced, tearing the nose landing gear from the aircraft and bending the propeller. The pilot and three passengers evacuated the aircraft without injury. The pilot reported that he suspected wind shear may have been a factor.

The Hughes 269 helicopter had completed a 100 hourly servicing during which, the main rotor abrasion strips were replaced, main rotor damper hydraulic fluid levels replenished, and the landing gear oleos checked to ensure that they were within limits. Ground and hover flight checks were then conducted to adjust blade weights and damper settings. The pilot and licensed aircraft maintenance engineer (LAME) reported that the wind was strong and gusting to about 30 kts. All the landings were made either crosswind or into-wind onto a hard bitumen surface. The LAME reported that, on each occasion, he made only small adjustments to the main rotor damper setting and blade track.

The pilot reported that on each of the flights prior to the accident flight, the helicopter had a significant vibration. At the end of each flight, the pilot would land the helicopter and disconnect the main rotor from the engine so that the LAME could make adjustments to the rotor head and blades. During the flights, the pilot occupied the right seat and the LAME operated the vibration and tracking equipment from the left seat. The LAME reported that during the accident flight, the helicopter's vibration levels and main rotor track had been adjusted to being well within limits, but soon after the pilot had made a very light and gentle landing, the helicopter entered ground resonance.

The pilot reported that after he landed the helicopter and as the engine speed decayed through approximately 2,500 to 2,300 RPM, the helicopter entered severe ground resonance. He forced the collective fully down and wound back the throttle setting. He also attempted to operate the rotor disengage switch but his attempts were hampered by the helicopter's severe vibration experienced during the ground resonance. A witness reported that about 5 seconds after landing, the helicopter rocked three times alternately on each of the two skids. It then spun through 360 degrees before disintegrating. Ground resonance occurs when unbalanced forces in the rotor system cause the helicopter to rock on the landing gear at or near its natural frequency.

Unless corrective action is taken, the amplitude of the vibration increases until the helicopter disintegrates. Corrective actions include immediately becoming airborne as ground resonance can only occur when the helicopter is in contact with the ground or stopping the main rotor as quickly as possible to remove the vibration source. After the helicopter had come to rest, the crew was unable to shut down the engine which continued to operate. A bystander reached into the cockpit to secure the engine and electrical equipment before he and others assisted the crew from the wreckage. There was no fire but the pilot was seriously injured. The LAME received minor injuries. No bystanders were injured.

An examination of the wreckage revealed that there appeared to be no pre-existing mechanical faults. The main rotor system vibration level and blade track were well within limits and the landing gear oleos had been checked during the servicing in accordance with the servicing manual. The landing gear oleos were checked after the accident and were found to be serviceable. Therefore, it was unlikely that a mechanical fault caused the helicopter to enter ground resonance. When the pilot reduced the main rotor speed prior to shutdown, the centrifugal force acting on the main rotor blades also reduced.

The main rotor blades would then have had a potential to excessively flap in response to the gusty wind. The helicopter may then have rocked on its landing gear oleos and subsequently entered ground resonance. Because the helicopter had no apparent mechanical faults, the prevailing wind conditions may have been a factor in the accident. The decision to perform the main rotor track and balance in strong and gusty wind conditions would appear questionable considering that the rotor system would have been experiencing varying degrees of translational lift.

Therefore, stable blade tracking and vibration readings from the equipment would have been difficult to obtain. Because the main rotor RPM was decreasing, the pilot was unlikely to have recovered the helicopter from ground resonance because both the recovery techniques were unavailable to him. He was unable to disconnect the main rotor from the engine due to the helicopter's vibration and, the quick development of the ground resonance meant that it was also unlikely that the pilot had sufficient time to re-accelerate the main rotor and take-off.