ferpe From France, joined Nov 2010, 2792 posts, RR: 59
Reply 4, posted (1 year 2 months 3 weeks 1 day 14 hours ago) and read 1661 times:
The failure in the QF31 case was from a mis-manufactured part, a counter bore was not center aligned and caused to thin material in an oilpipe feeding into the rear bearing for the IP shaft to fatigue and break. The oil that sprayed out caught fire and weakened the highly loaded IP disc, it came loose from the shaft, over-speeded and disintegrated. So the primary action was to find all oil pipes that were not correctly made and change them, this was done with a boroscope action on all engines.
At the same time it became clear that there were 3 different revisions of the rear IP bearing structure, Quantas has an extra high revving version of the engine for more thrust (for their SYD-LAX legs). This apparently loads this bearing structure so that they need the C revision to have acceptable time on wing. As QF and many others are on power by the hour RR can swap engines so that QF gets the sturdier ones while others can do with e.g. A or B rev ones.