Characterization of pitting damage and prediction of remaining fatigue life

Quantification of corrosion and modeling of subsequent fatigue crack growth have been the focus of much intense inquiry in recent years. In previous work, the present authors proposed a crack growth model for this fatigue damage scenario that incorporated recent laboratory observations and built upon the approaches of other researchers. Here this model is extended to establish fatigue life prediction distributions for pre-corroded 2024-T3 aluminum using random cross-sectional plane micrographs as the damage quantification means. First, using automated image analysis, metrics are extracted from these micrographs that are shown to bear relevance to the fatigue performance of the material. Insight was gained by studying this pitting from a number of different perspectives, including analysis of spatial distribution and morphology. Based on the observation that the failure of individual, pre-corroded, fatigue specimens often results from the simultaneous nucleation of several cracks, a life prediction method is presented. Here, multiple crack evolution scenarios are simulated with the understanding that the distribution of lowest resulting lives will be limiting and thus representative of the experimentally observed lives. The resulting life predictions are shown to compare well against experimental values. Model sensitivities are addressed and the merits and limitations of the proposed methodology are discussed.