Formal Report AAR 7/2010. Report on the accident to Aerospatiale (Eurocopter) AS 332L Super Puma, G-PUMI at Aberdeen Airport, Scotland on 13 October 2006.
The accident was notified to the Air Accidents Investigation Branch (AAIB) by the Operator’s Flight Safety Officer. The following Inspectors participated in the investigation:
Mr R D G Carter Investigator-in-charge Mr C A Protheroe Engineering Miss G M Dean Operations Mr P Wivell Flight Recorders
The aircraft was departing from Runway 14 for a flight to oil platforms in the North Sea, carrying 13 passengers. Five seconds into the takeoff the crew heard a bang and an abnormal vibration started. The crew rejected the takeoff and landed back on the runway. The aircraft started to taxi but the severe vibration continued so the commander stopped and shut down the helicopter on the threshold of Runway 32.
Initial examination showed that one main rotor blade spindle had fractured, through the lower section of its attachment yoke on the leading side of the spindle. Post-fracture plastic deformation of the lug had stretched open the fracture, separating the faces by some 12 mm.
As a result of this accident the helicopter manufacturer published an Emergency Alert Service Bulletin, requiring periodic inspections, and this was subsequently mandated by the European Aviation Safety Agency (EASA) as an Airworthiness Directive. In July 2009 the manufacturer issued Service Bulletins which introduced a ‘wet’ assembly procedure, with new nuts, for the main rotor blade spindles. This eliminated the requirement for the repetitive inspection procedure and was made mandatory by the issue of an Airworthiness Directive (AD) by the EASA.
The investigation identified the following causal factors for the failure of the spindle yoke:
(i) Wear on the flapping hinge inner race.
(ii) Excessive clamping pre-load across the yoke, due to the tie bolt being torqued to the specified dry value in the presence of grease when it was reinstalled some 175 hours prior to failure of the yoke.
(iii) Significant hoop stresses in the bore of the yoke due to adverse tolerance stacking and the associated interference fit of the bush in the yoke.
The following were considered as contributory factors in the failure:
(i) Flight loads biased towards the high-speed level flight condition, slightly higher than those generated by normal level flight cruise conditions.
(ii) A minor deviation in corner radius profile at the inner end of the bore of the yoke, with a small increase in the attendant stress concentration.
(iii) A minor reduction, at the fatigue origin site, in the intensity of the compressive surface layer stresses from the shot-peen process.
(iv) Flight loads in the spindle yoke slightly higher than anticipated in certification fatigue testing, due to the action of the lead-lag dampers (frequency adaptors).
One Safety Recommendation is made, to the EASA, concerning HUMS detection in helicopter rotating systems.
Download full report:
AAR 7-2010 G-PUMI.pdf (2,888.23 kb)