History of Flight
On 6 August 2003, a private pilot flew an Enstrom F-28F Falcon (registration G-BXXW) with his wife and child to stay with friends near Wells, Somerset. The helicopter landed on a village cricket pitch and remained overnight. The following day, the pilot intended to return to Sheffield with a stop at Wellesbourne Mountford Airfield.
Before departure, the pilot conducted a pre-flight check, including the main rotor gearbox (MRGB) oil level. He described this check as awkward due to the sight glass position and the need for a torch. The oil level appeared normal, and no oil spillage or seepage was observed. The engine required an additional pint of oil. The rest of the checks revealed no issues.
The helicopter departed at approximately 1415 hrs, climbing to 1,500 feet and 85 mph IAS. Passing RAF Lyneham toward Kemble, the pilot requested a radar service from RAF Brize Norton, but it was unavailable. He was instructed to remain clear of the Brize Norton CTR.
About 10 nautical miles south of Wellesbourne, as the pilot prepared to contact the airfield, a vibration increase occurred, followed by a 'thump' and a yaw to the right. A burning smell entered the cockpit. The low rotor RPM warning horn sounded; the rotor RPM needle dropped below the green arc while engine RPM rose above. The pilot fully lowered the collective, applied full right yaw pedal, raised the nose to reduce speed to 65 mph, and closed the throttle slightly. He made a MAYDAY call to Wellesbourne, reporting the position and engine failure.
The helicopter was over a rough grass field with steeply sloping ground. The pilot turned 180° to stay over the field and into wind, maintaining full down collective and full right yaw pedal. At about 50 feet AGL, he flared, raised the collective to cushion the touchdown, and centralised the pedals. The helicopter touched down gently with no forward speed. After landing, the pilot informed Wellesbourne of the safe landing.
As the family disembarked, a small fire was seen under the gearbox fairing. The pilot shut down the fuel and extinguished the fire with the aircraft's extinguisher. The gearbox drive belt pulley was glowing red-hot. Fire services arrived about ten minutes later and remained until the risk of fire passed. The pilot later observed oil leaking from the helicopter's underside.
Engineering Investigation
Damage included twisting and bending of the tail rotor drive shaft due to severe shaft 'whirling'. The shaft impacted and damaged the top of the tail boom, distorted the laminates of the Thomas coupling at the tail rotor gearbox input, and sheared the input shear pin. The MRGB input drive shaft fractured; its belt drive pulley moved 1-2 inches rearward, damaging a Thomas coupling at the forward end of the tail rotor drive shaft. The drive belt was displaced rearward, and the belt tensioner was damaged. A small fire fueled by MRGB oil occurred. Only about ¼ pint of blackened oil could be drained from the MRGB. The chip detector was blackened with oily sludge, with no visible chips.
At the rear bearing, where the shaft fracture occurred, the bearing cover bolts had lost torque, and the gasket showed evidence of a small oil leak. The lower skins were covered in oil film. The self-aligning double row barrel roller bearing had broken up, with some rollers jammed sideways, causing the bearing to seize. This generated considerable heat, softening the shaft and possibly causing loss of torque.
The MRGB was stripped and examined. Chips were found on the magnetic probe of the gearbox rotor speed sensor. The rear bearing had extensive wear, and the bearing cage failed due to fatigue. Temperatures exceeded 650°C at the inner race. The forward bearing was undamaged. Lubrication of the rear bearing is more restricted than the forward bearing; it relies on unpressurised flow through a small oilway. No evidence of oilway restriction was found. The report concluded that wear occurred between the rear bank of barrel rollers due to a breakdown in lubrication, leading to cage failure and bearing seizure. Heat softened the case hardening on the pinion shaft, reducing fatigue resistance, and slackness likely caused multiple fatigue cracks, finally fracturing the softened shaft in torsion.
Chip Detector Wiring History
The AAIB learned from the Enstrom agent that the helicopter had a history of MRGB chip detector warning indications. After the accident, the chip detector wiring was found to use an 'automotive type' crimp fitted over the wire insulation. The pilot stated he had fitted the crimp after the chip detector wiring broke while he was checking it. He obtained parts from a local electrical spares outlet and crimped the wires after baring the conductor. He also supplied a new chip detector to the agent for installation when new rotor blades were fitted, assuming the wiring would be checked. At that time, only a visual check of the chip detector and a 'press to test' of the warning light were required; no continuity check was mandated. After this accident, Enstrom introduced a Maintenance Manual change to include a continuity check of chip detector wiring for F-28 and 280FX series helicopters.
Analysis
The investigation determined that the rear bearing of the MRGB seized due to inadequate lubrication. The limited blackened oil supported this conclusion. The rear bearing is more adversely affected by reduced lubrication because of its restricted lubrication path. Bearing seizure caused pinion shaft fracture, loss of tail rotor drive, and damage to the tail boom and transmission.
The reduced lubrication was likely due to a low oil level in the MRGB, although the reason could not be definitively established. A possible pre-existing leak at the rear bearing cover gasket would have been a weep or drip. The subsequent shaft fracture allowed a large oil leak, which the pilot observed after landing. The difficulty of reading the oil level through the sight glass may have contributed to an undetected low oil level; a meniscus might give a false normal appearance. Alternatively, even with adequate oil, a pre-existing bearing defect or lubrication path blockage could explain the failure, but no evidence of such was found.