I would say that SLL is a simplistic concept in that it measures time only. Fatigue crack propagation is also a function of load level with each cycle. I would hope that SLL is set with very conservative assumptions on the spectrum of load levels. Otherwise you would need to consider not only Nr but simultaneous Torque during all running. Even then, fatigue damage done in early life can be more significant than later. This was the case IIRC with tests on the bevel gear shafts.
Last edited by Concentric; 17th Dec 2016 at 10:28.