Probability modeling and material microstructure applied to corrosion and fatigue of aluminum and steel alloys

Engineering prognosis and life cycle management have increasingly become dependent on probabilistic methods and material microstructure. Due to variability in microstructural properties produced by differences in alloys and processing conditions, scientifically based probability modeling for life prediction and reliability is a challenging problem. For several years, probabilistic and statistical methodologies have been available to address prognosis issues, and there has been significant progress in computational methods for engineering applications. Their utility has been greatly limited by the lack of well developed physically and mechanistically based models for many of the complex applications. Without a reasonably accurate description of the damage process at a scale that is pertinent to the desired application, probabilistic computations have minimal value for prognosis and life cycle assessment. In this paper, the major methods in computational probability are reviewed, and their efficacy is illustrated through two engineering examples. One example is corrosion and corrosion fatigue crack growth for 7075-T6 aluminum alloy, which is used commonly for structural components in commercial and military aircraft. The other example is very high cycle S–N behavior of SUJ2 bearing steel for which multiple modes of failure are manifest in the high cycle regime. Their application to other critical engineering problems is discussed.