Forced/precautionary landing

Survival

The pilot survived the crash but suffered multiple injuries. He undid his safety belt on the right front seat and exited the helicopter from the damaged left side. However, he died before the wreckage was located. The accident was potentially survivable. The pilot carried his flying helmet in the helicopter but had elected not to wear it for the ferry flight. He suffered serious head injuries.

Pilot information

The pilot held a Canadian commercial helicopter pilot licence. He was endorsed on the Bell 205. His total flight time was 10,092 hours, of which 9,992 had been flown in helicopters. He had flown approximately 3,600 hours in the Bell 205. His last company flight cheek was on 13 June 1995. He passed his last aviation medical examination on 7 November 1995. He was experienced in carrying out helicopter fire-fighting operations.

Medical information

The pilot's medical certificate required the use of vision correction lenses when flying. Evidence at the accident site indicated that he was probably wearing spectacles at the time of the accident. The only physiological condition subsequently found, which may have slightly affected his flying performance, was due to skin irritation, after being doused with jet fuel while refuelling just before departing Cockatoo.

Damage to aircraft

The helicopter was destroyed during the impact sequence. There was no post-impact fire.

Weight and balance

The helicopter was within its approved weight-and-balance limits at the time of the accident.

Weather

The wind was reported to be north-westerly at about 5 kts. The cloud was one or two oktas of stratocumulus at 5,000 ft. Visibility was in excess of 10 km. The weather was not a factor in the accident.

Aids to navigation

The helicopter was equipped with a global positioning system (GPS) which would have allowed the pilot to accurately fly a direct track to Benalla.

Wreckage information

No fault was found with the airframe or electrical systems of the helicopter. On-site examination of the wreckage indicated a loss of drive from the engine to the main transmission. The core engine was found to be physically disconnected internally from the main engine output gearbox. The engine chip detector was found to be heavily contaminated with metallic debris. A subsequent engine teardown inspection identified that the engine output gearbox had suffered significant damage and was the source of the drive failure. A discontinuity found in the engine chip detector wiring was assessed as impact damage.

Maintenance summary

Maintenance history was obtained from journey logs, engine logs and Transport Canada Aircraft Technical logs (engine and airframe). The following is a summary of maintenance data relevant to the engine fitted to GFHO at the time of the accident:

• 04 Apr 1992; engine, serial number LE-07683C, upgraded from T53-13B to T53-17A configuration at 6,624 hours time since new (TSN).
• 16 May 1994; engine LE-07683C overhaul completed at 7,020.4 hours TSN.
• 17 Aug 1994; engine LE-07683C fitted to GFHO.
• 05 Sep 1995; governor input seal leaking. Governor assembly replaced with overhauled unit.
• 08 Sep 1995; GFHO last flight in Canada.
• 25 Sep 1995; 100 hr inspection complied with in accordance with Bell 205 M & 0 and Lycoming manuals. Engine chip plug contains small amount of fuzz. All remaining screens and detectors found free and clean of debris.
• 29 Sep 1995; engine LE-07683C accessory gearbox, serial no. 4029 replaced when copper-coloured metal contamination was found in the airframe paper filter. Accessory gearbox, serial no. 5154-6 (17.5 hours time since overhaul) fitted in accordance with Lycoming MM. Ground run & leak check complied with.
• 04 Oct 1995; helicopter GFHO disassembled and shipped to Australia.
• 24 Nov 1995 (late entry 19 Dec 95); engine LE-07683C, N2 torquemeter boost pump, serial no. Cl 198 replaced when "oil not scavenging from N2 gearbox and brass found in engine filters". N2 torquemeter boost pump, serial no. LA7278 fitted. "Pump removed has internal damage to brass bushings and the shaft has moved forward and uncoupled from pump".
• 28 Nov 1995; helicopter GFHO reassembly completed in Australia.
• 29 Nov 1995; GFHO first flight in Australia.
• 13 Dec 1995; at the time of the accident, engine time logged in Australia approximately 4.5 hours, engine had accumulated approximately 7,413.0 hours TSN and therefore 392 hours time since overhaul (TSO).

Post accident maintenance observations

• Oct 1995; according to the operator, the accessory gearbox removed on 29 Sep 1995 was disassembled and found to be acceptable for continued use and not responsible for the oil contamination found in the filter. It was decided that the N2 torquemeter boost pump may have caused the metal contamination and that this should be replaced when the aircraft was re-assembled in Australia.
• 24 Nov 1995; the torquemeter boost pump, as well as suspected to be the source of metal contamination, was considered to be responsible for a previous oil leak. According to the operator, the oil leak mentioned was discovered during the governor change on 5 Sep 1995. Subsequent inspection of the replaced pump found incorrect assembly had led to gouging of a brass bushing and generation of brass contamination.

Engine teardown inspection

An engine teardown inspection revealed that the helical sun gearshaft was uncoupled from the output reduction gearbox. The helical sun gear is coupled to the power shaft and drives the reduction gearing. Output power is extracted from the reduction gearing through an externally splined output shaft. The uncoupling resulted in the loss of drive from the engine to the main transmission. Disassembly also revealed:

• Copper/brass and steel particles contaminated the oil system.
• The helical sun gear (PN 1-030-192-04) had apparently overheated and all its teeth had been machined off by the three mating planet gears mounted on the reduction gearbox carrier housing assembly. The input gear teeth of the planet gears were extensively damaged, the involutes being filled with smeared metal. Two helical sun gear teeth appeared to have fractured as a result of fatigue crack growth prior to the general destruction of the gear. A gear tooth fragment, matching the remnant of the helical sun gear tooth fracture, was found in the coarse filter screen of the oil pump.
• The N2 tachometer drive spur gear (PN 1-070-062-04) was loose on its shaft. The nut (PN MS 172237) and nut retainer (cup washer) (PN 1-070-066-01) securing the gear to the shaft were missing from the assembly. The nut was found loose in the accessory drive carrier assembly. The nut retainer, used to lock the nut on the shaft, was found in the scavenge strainer screen of the accessory drive gearbox, with a broken locking tang. Movement of the spur gear resulted in impact and damage to the drive spur gear teeth on the power shaft, reduction gearbox outer housing and No 21 bearing clamping plate.
• The sleeve bushing of the N2 tachometer drive gear assembly had moved out of its housing and the lower bearing cage had failed. The sleeve bushing retention pin (PN AN122683) missing from the assembly was located in the metal debris subsequent to the teardown inspection. The retention pin hole on the sleeve bushing was not properly located with respect to depth and location.
• The engine oil pump, mounted on the accessory drive gearbox had seized. The input drive shaft had sheared. A metal sliver (5mm x 2.5mm) jamming a pump impeller blade had caused the seizure. The chip detector was completely covered in particles.
• Extensive metallic debris was found in the accessory drive gearbox assembly, however, the drive gears and bearings were undamaged. The scavenge strainer screen was blocked with metal debris. Some of the debris was identified as a cup washer from the N2 tachometer drive spur gearshaft and a helical sun gear tooth.
• The sun gearshaft roller (No. 21) bearing (PN 1-300-082-03M) had completely failed, with no roller elements remaining. A visual examination assessed the failure to be as a result of metal contamination and oil starvation. The power shaft acting as the bearing inner race had been damaged as a result of excessive heat and the skidding rollers of the bearing.
• The forward compressor ball bearing and aft compressor roller bearing, mounted on the power shaft, suffered damage consistent with oil starvation.
• The torquemeter boost pump fitted to the engine at the time of the accident was found to be undamaged, correctly assembled and free to rotate.

Reduction gearbox carrier housing assembly inspection

The sun helical and planetary gear is located in the reduction gearbox carrier housing assembly (PN 1-030-340-04). Examination of the GFHO carrier housing assembly following the accident revealed that:

• The roundness (degree of ovality) of the aft bearing bores was up to 8 times greater than the allowable tolerance of 0.001 inches. The concentricity or alignment of the small bearing (aft) bores with the large bearing (front) bores was found to be up to 25 times greater than the allowable tolerance of 0.001 inches.
• The carrier housing large bearing (forward) bores were within tolerances.
• The lack of parallelism between the front and rear plates was found to be up to 7 times greater than the allowable tolerance of 0.001 inches.
• The bores displayed evidence of fretting damage.
• The carrier assembly did not have a vendor manufacturing code (required for civil certified components) etched into its surface.

Engine maintenance records recovered by the Transportation Safety Board of Canada, indicated that repairs at the last engine overhaul included the chrome plating and grinding of the six bearing bores of the carrier housing assembly in accordance with the T53 overhaul manual. The concentricity and roundness of the front and rear bore holes of the carrier housing assembly were within limits at the time of the overhaul.

ELT and mobile telephone

After the accident both ELTs carried in the aircraft were found to be switched off. The Narco ELTIO which should have been selected to ARM in compliance with company policy, had the capacity to be activated by crash impact if ARM had been selected. Both survival beacons were found to be capable of normal operation after the accident.

The pilot had access to a mobile telephone fitted to the helicopter. However, even if he had been able to reach it he would not have been able to use it to alert authorities of the accident because the battery had separated from the telephone during the accident. When a serviceable battery was subsequently fitted, the telephone operated normally. The telephone installation was also connected to the aircraft power supply. Inspection of the aircraft indicated the power supply plug had been pulled from the telephone in the accident. Also, severe impact damage to the helicopter's nose area had severed one of the aircraft electrical cables at its point of attachment to the battery relay.

Flight following

Two very high frequency (VHF) radios were fitted to the aircraft for flight following with air traffic services. However, the pilot elected not to use these radios for monitoring of the flight by Airservices Australia. Instead he contacted the DCNR airdesk radio operator and advised of his departure time. The radio he used for the call was provided by DCNR to allow communications on the Victorian public sector mobile radio network (SMR). This equipment was not compatible with the aviation communications network. The pilot used the "trunked" function of the SMR equipment, whereby communications could only be heard by the person transmitting and the person at the specific station the pilot was calling.

The airdesk operator provided "flight following" for the flight. Under the DCNR monitoring system, the pilot was required to make radio contact with the operator at least once every 30 minutes. Following the departure call from Cockatoo, the airdesk operator received no further calls from the pilot. At 1720, the airdesk operator called the helicopter by radio and also by telephone but received no response. At 1725 the airdesk operator contacted the Benalla rappelling crew which was expected to train with the helicopter in the next day or so. A crewman advised that he had called the pilot on the trunked radio at 1636, to ask when the helicopter was expected to arrive at Benalla. The pilot's response was that he would arrive in about 35 minutes (1711 ESUT).

Search and rescue

At 1736 the DCNR airdesk operator contacted Melbourne Flight Service to establish if the pilot of GFHO had transmitted any flight details to Airservices Australia. The Flight Service operator advised that no radio calls had been received from the pilot. At 1802 the DCNR operator advised Melbourne Flight Service that the helicopter was missing. The Flight Service operator in turn relayed the information to the Melbourne Search and Rescue officer. The uncertainty phase of search-and-rescue (SAR) procedures was activated at 1806. Unsuccessful checks to locate the helicopter were made by SAR staff. At 1840 the alert phase was activated. At 1919 the distress phase was activated, 2 hours and 8 minutes after the pilot's estimated time of arrival at Benalla. Search activities continued throughout the night. By the morning of 14 December a large-scale search was under way with 24 helicopters and seven fixed wing aircraft used in the search.

By 1822, on the previous day, the Australian company operating the helicopter had dispatched a helicopter to search for the missing Bell 205. The search pilot estimated that GFHO could have been somewhere in the Marysville area when 35 minutes flight time from Benalla, so he tracked direct to Marysville to commence the search and continued searching while monitoring the ELT distress frequency until 2048, last light being 2101. Marysville is 5.5 km SE of the accident site.

Search co-ordinators and search aircraft focused efforts in the early stage of the search on identifying the source of transmission signals on the distress frequency in the Strathbogie area, about 61 km N of Buxton.

Several people had seen and heard the helicopter minutes before the crash. A couple had heard what was, in hindsight, probably the sound of the helicopter crashing. These people did not hear or see enough to convince them that an accident had occurred. However, they listened to or watched the evening news and when nothing was mentioned about a helicopter accident or a missing helicopter, they did not contact the police. The search authorities made no public media release of the missing helicopter until about 0630 on the morning after the accident. As a result of the media release, police received the first of several public reports of sightings of GFHO between Buxton and Narbethong on the previous evening. Narbethong is 13 km SSW of the accident site. At 0645 police search-and-rescue officers dispatched two units to the Buxton area. At 0916 a police helicopter crew spotted the wreckage of GFHO while searching an area of reported sightings near Buxton the previous day. Most of the sightings were reported to police as a result of the media release. The accident site was 5 km right of the direct track from Cockatoo to Benalla.

Autorotation

Autorotation is the means by which a pilot may safely land a helicopter in the event of no engine power driving the rotors. In single-engine helicopters, loss of drive to the main rotor normally occurs as a result of engine failure. When the engine gearbox failed in GFHO, the engine power output to the main rotor transmission was effectively uncoupled but the engine possibly did not stop immediately. Loss of drive to the main rotors, excluding engine failure, is a very rare event necessitating action by the pilot to effect entry into autorotation in order to maintain rotor RPM. The gearbox failure resulted in metal debris seizing the scavenge oil pump causing overload shearing of the oil pump input shaft; this resulted in sudden loss of engine oil pressure, rapid engine overheating and power loss.

According to the approved flight manual for the Bell 205, the optimum airspeed for an autorotative descent, at a gross weight above 7,500 lb, is 55 to 60 kts. In the case of GFHO, the rate of descent in a stabilised autorotation would have been about 1,900 ft/min. At 35-45 ft above the ground (or in this case above the trees), the pilot flares the helicopter which decreases both the rate of descent and the forward airspeed, followed by levelling the helicopter and increasing collective pitch at about 4 ft. Correct autorotative technique ensures that the helicopter arrives on the ground or contacts the tops of the trees with virtually no rate of descent and very low forward airspeed. For an autorotation into the trees, zero airspeed is preferred at the top of the trees. The ground distance covered from the moment of engine failure to entering a stabilised autorotation to touchdown varies according to the pilot's time to assess and react, airspeed at entry, gross weight of the helicopter, temperature, wind velocity and, in particular, the height above ground or tree tops when the autorotation commences. If the height is greater, so is the potential range.

Evidence at the accident site indicated that GFHO probably had a descent rate of about 500 ft/min and a forward airspeed up to 50 kts when it impacted the trees.

1. Misalignment between the helical sun gear and its planetary gears resulted in accelerated wear, high loading and a reduction in fatigue life of the sun helical gear.
2. Two helical sun gear teeth failed due to fatigue crack growth.
3. Gear and metal debris lodged in the engine oil pump and seized the pump resulting in loss of engine oil pressure.
4. Drive to the output gearbox was lost, forcing the pilot to perform an emergency landing.
5. By not arming the ELT, the pilot diminished his chances of being found quickly after the accident
6. Vital information was not reported to search co-ordinators for many hours because of a delayed public media release about the missing helicopter.

Refuelling spill

During motorised refuelling at Cockatoo, a hose coupling separated, causing jet fuel to spray over the pilot, dampening his clothes. The pilot was seen to wash his hands and face with water before take-off for Benalla. With jet fuel on his clothes, he would have suffered a mild but worsening degree of skin irritation in flight. As the pilot had only flown about 59 km from Cockatoo to the accident site, his flight time would have been about 20 minutes. From the time of refuelling, the time lapse was probably about 30 minutes, by which time he might have been experiencing an uncomfortable burning sensation on his skin. Despite the jet fuel irritation, the pilot's ability to perform an autorotation should not have been impaired significantly.

Engine failure analysis

The Transportation Safety Board of Canada, Transport Canada, and the engine manufacturer assisted in the investigation. The Bureau based the engine failure analysis on the evidence and information presented by these agencies and other parties which included investigation research and analysis in Canada and the United States. The Bureau's findings are based on the most probable engine failure sequence drawn from information received during the investigation.

Helical sun gear

Examination of the remnants of the helical sun gear teeth revealed that two teeth had fractured as a result of fatigue crack growth, prior to the general destruction of the gear. A gear tooth fragment, matching a remnant, was found in the coarse filter screen of the oil pump. A specialist report stated "An examination of the fracture surface revealed that the separation of the gear tooth was caused by fatigue crack growth. The features of the fracture indicate that fatigue cracking initiated in the root radius on the driven side of the tooth. Uniform spalling and the development of cracks in the surface of gear teeth are caused by the high alternating contact pressures created during the meshing of the gears transmitting high loads. The recovered section of the gear tooth showed no evidence of abnormal heating. The general condition of the section of gear tooth is consistent with operation in the presence of lubrication." This indicated that the fatigue failure of the helical sun gear teeth preceded the oil pump failure. The overheat condition was considered to be secondary damage caused by loss of lubrication.

Reduction gearbox carrier housing assembly

Examination of the carrier housing assembly revealed large out-of-tolerance ovality and concentricity between the front and rear bearing bores. At the request of BASI, and under its direction, the manufacturer analysed the effects of a planetary gearshaft misalignment to address bore mis-machining. The manufacturer concluded that the time to failure correlated with the level of gear mesh misalignment found.

Repairs to the six bearing bores of the carrier housing assembly were carried out at the last engine overhaul. Examination of documentation relating to this procedure indicated that the concentricity and roundness of the front and rear bore holes of the carrier housing assembly were within limits following the repair. The errors in ovality, concentricity and parallelism found after the accident were therefore considered to be as a result of forces produced during the engine failure sequence and during disassembly of the engine following the accident.

The engine manufacturer examined the documentation relating to this procedure and considered that bearing bore true position errors may have been introduced during the grind/plate process as a result of the tooling fixture that was used.

The engine manufacturer confirmed that the carrier housing assembly in GFHO was not approved for use in civilian T53 engines. However, this was not considered to have contributed to the failure of the engine.

N2 tachometer drive spur gear assembly

The locking tang of the nut retainer on the assembly was found to have fractured due to overload. This may have occurred during assembly of the spur gear at the last overhaul. Wear marks were evident around the indentation that locked the nut retainer onto the nut. Cup washer failures are known and the manufacturer has since addressed the assembly procedures.

Oil pump

The metal jamming the oil pump was identified as case hardened material, indicating that it was gear material. The size and shape of the material suggested that it was a chip of a gear tooth.

Engine drive failure

BASI was unable to satisfactorily determine the primary event leading to the engine failure. From the information available, two possible engine-drive failure sequences were identified:

Sequence 1

• The locking tang of the nut retainer on the N2 tachometer drive assembly fractured due to overload, possibly during assembly at the last engine overhaul.
• The nut on the gearshaft lost torque during engine operation and gradually backed off the shaft with the cup washer. The nut and cup washer fell into the inlet housing.
• The loose N2 tachometer drive spur gear resulted in wear on the threads of the gearshaft, damage of the reduction gearbox outer housing and damage of the No. 21 bearing clamping plate.
• Debris from the wear and damage of the components entered the No. 21 bearing. Damage to the bearing resulted in misalignment of the sun gearshaft with its mating planetary gears.

Sequence 2

• Repairs to the reduction carrier assembly bearing bores at the last engine overhaul resulted in planet gear bearing bore true position errors. This resulted in misalignment of the sun gearshaft with its mating planetary gears. Vibratory forces resulted in the backing off of the N2 tachometer drive nut and cup washer from its shaft.

Common sequence

• Misalignment between the planetary and sun gear resulted in accelerated wear, high loading and a reduction in fatigue life of the sun helical gear. Two gear teeth from the sun helical gear fractured due to fatigue crack growth resulting from the abnormal loads.
• Abnormal loading resulting from the misalignment of the loose N2 tachometer drive spur gear caused the bearing sleeve to work out of the accessory carrier and damage to gear teeth on both the N2 spur gears.
• A tooth chip from the sun helical gear or the N2 drive spur gear was carried with scavenge oil and lodged in the engine oil pump, jamming an impeller. This caused the overload failure of the oil pump input drive shaft, halting oil supply and scavenging.
• Loss of oil pressure resulted in high temperature damage to the forward and aft compressor bearings and No. 21 bearings, and to the reduction gearbox carrier assembly gears and bearings.
• Lack of lubrication and misalignment of the sun helical gear resulted in the complete stripping of its gear teeth.
• The abnormal loading on the gear assembly deformed the rear reduction gearbox carrier and its bearing bores.
• Failure of the sun helical gear resulted in loss of drive to the output gearbox and no engine power being transferred to the main transmission.

Lubrication system contamination troubleshooting

On 29 Sep 1995, it was reported in the journey log that the airframe paper filter contained copper-coloured metal contamination. According to the maintenance organisation, they suspected a problem with the accessory gearbox and replaced it at this time. Additionally, following discussions with the engine repair shop it was assessed by the maintainer that the N2 torquemeter boost pump could also have been responsible for the brass contamination found in the engine filter. It was decided that the torquemeter boost pump would be changed following shipment to Australia.

Subsequent inspection of the removed accessory gearbox indicated that it was not responsible for the oil contamination. The torquemeter boost pump was reportedly replaced prior to commencement of flying operations in Australia and when inspected was found to have been mis-assembled. This had resulted in loss of drive to the boost pump rotors and generation of metal from a brass bearing bush. It was considered that the brass in the engine filter was from the brass bearing. The operator provided this information after the accident and at that time it was not entered in the aircraft maintenance logbooks.

Maintenance records did not report that the manufacturer's recommended lubrication system contamination troubleshooting procedures as detailed in the T53-17A maintenance manual were followed. The metal found in the engine filters may have been an early warning that the engine failure was imminent. The engine manufacturer believed that if the maintenance manual procedures had been followed, the engine could have been flagged for repair or possible removal prior to the accident.

The operator considered that the intent of the requirements in the engine manufacturer's troubleshooting procedures was carried out. Through systematic checks and inspections the source of the brass metal contamination was determined to be the torquemeter boost pump. By replacing the defective part and carrying out ground runs the operator considered that standard procedures were followed.

The failure sequence may have occurred over such a short period of time, that there was little forewarning of the impending engine failure. No reports of chip lights illuminating were entered in the maintenance logs following the aircraft's arrival in Australia.

Autorotation

The reason the helicopter retained substantial forward speed and rate of descent when it impacted the treetops could not be determined.

Search and rescue

The fact that the ELTIO was off and not switched to the preferred selection of ARM, seriously diminished the pilot's chances of being found quickly after the accident.

Police found the accident site largely in response to witness sighting reports received after the media release about the missing helicopter. The delay in the media release may have diminished the pilot's chances of survival.

The Canadian registered helicopter C-GFHO, along with a pilot and maintenance engineer, had recently been shipped to Moorabbin Airport from Canada. It had been partially dismantled for transport to Australia and was re-assembled at Moorabbin Airport during November 1995. It was to have been part of the fire-fighting service to be operated during the fire season by the Victorian Department of Conservation and Natural Resources (DCNR). For the fire-fighting role, the helicopter had the capability to ferry fire crews and to be used for rappelling operations. It was also equipped with a belly tank for fire bombing. DCNR had planned to base the helicopter at Benalla for the fire season. Some initial testing and training had been done with the aircraft after re-assembly, but it had not formally started fire-fighting operations.

Early on the afternoon of 13 December 1995, the pilot ferried the aircraft from Moorabbin Airport to Cockatoo, 35 km ENE of Moorabbin. During the afternoon it was used at Cockatoo to assist in the training of volunteer fire fighters. During this training the pilot gave familiarisation briefings to the volunteers, including demonstrating the helicopter emergency locator transmitters (ELTs). Two were carried; one was a marine survival beacon, and the other was a Narco ELTIO designed for aviation use. Subsequently, during start-up for a demonstration flight, the pilot discovered that the ELTIO was inadvertently switched on and transmitting the distress signal. He turned the ELT off but did not arm it for impact activation.

On completion of the training exercise, the pilot departed Cockatoo at 1623 ESUT for Benalla, to position for the start of operations. About 5 km S of Buxton, 105 km SSW of Benalla, the engine gearbox failed resulting in loss of drive from the engine to the main transmission. The pilot performed an autorotative descent but failed to reach a clearing. After initial tree impact, the helicopter travelled 60 m horizontally before ground impact. The trees were approximately 25 m high and the terrain was undulating. The helicopter was destroyed in the accident.

The wreckage was difficult to see from the air because of the tree canopy. Despite a subsequent search, the helicopter was not located until about 0916 ESUT on 14 December, approximately 16.5 hours after the accident.