Reliability-based cumulative fatigue damage assessment in crack initiation

A reliability–based fatigue life assessment of notched specimens under variable amplitude loading is presented. The proposed methodology adopts the Monte Carlo simulations and the local strain life approach to take into account the effects of loading variability, material properties, modelling errors, and cumulative damage variations on fatigue life estimates. Statistical analyses were conducted to model the statistical distribution for every random variable. The results of the SAE Cumulative Damage test program were validated with three different types of SAE random loading histories applied to the test specimens made of a RQC–100 steel. It was shown that the cumulative fatigue damage data followed the normal distribution but the mean and COV varied with different life prediction techniques. Moreover, the important sources causing uncertainties in life predictions were identified.