Bearings
Additionally, knowing that the Ball Bearings act as Thrust bearings and Axial bearings under load, we can describe the "Limit" Sphere as certainly Three Dimensional. The Balls operate in a "Tube", and react against their race depending on loading. If there were only thrust loads to attenuate, the Wear Print would be at the "Forward and Aft" corners of each metal run. A total of four separate areas. If Radial loads present, these four areas would expand, as the force of the Shaft expresses itself in radial directions as well as in axial stress. If these loads are balanced, (reasonably), there could be an expectation of this design working through and beyond its service life.
This is not the case. Mechanical loads vary widely depending on several designed for aspects: Axial shift depending on translation of Thrust, RPM and lubrication, plus vibration, put variable and tested for loads on this bearing.
As above, if the individual balls are demanded to snub stress in three dimensions, the wear print will eventually affect the entire face of the Race. Under constant load in three dimensions, Oil can be wiped, and metal to metal can accelerate wear. The "C" mod AD requires an initial inspection at 200 cycles, then the next at 100 cycles after it. Why?? As installed, the 'new' bearing is at minimal tolerances. With use, the tolerances widen. As before, I describe a logarithmic failure graph with wear. The more it has worn, the faster (rate) it wears. 200 cycles is about a year, 100 about six months.
Inspections are ordered to find wear. That means, wear is expected to be found. Initially, wear is less concerning, but with sloppier tolerances, inspections are demanded more often. Eventually, the bearing will require rebuilding, or replacing with new. How this affects the original Certificate is between RR and the operators and regulators. And the travelling Public.
So the "A" of the abcde failure cascade is Duff bearing. Not support structure, not oil pipe, not EEC, not Oil Fire. Think..... The Oil Fire caused Aft migration of the IP Shaft? No.
Fire is an instant and chaotic failure, but it cannot wear out bearings and Splines in mere seconds. Especially one meter forward.