LOW CYCLE FATIGUE: PROBABILITY AND STATISTICAL MODELING OF FATIGUE LIFE

Low cycle fatigue (LCF) induces damage accumulation in structural components used in various applications. LCF typically describes conditions for which plastic strains are larger than elastic strains. In order to certify and qualify a structural component, manufactured from a given material, that requires high reliability for operation and safety, fundamental material properties should be experimentally investigated and validated. The traditional strain life approach serves as the underlying experimental method for most LCF investigations. Building upon that background, the purpose of this paper is to investigate the statistical variability and appropriately model that variability for life in LCF. Specifically, the variability associated with the median behavior in a strain life graph for data is examined. The ensuing analyses are based on data for a cold-rolled, low carbon, extra deep drawing steel; ASTM A969 which is appropriate for applications where extremely severe drawing or forming is envisioned. It is frequently used in the automotive industry for components such as inner door components and side body components. For substantiation of the proposed modeling techniques, data for 9Cr-1Mo steel is also investigated. Such steel is frequently used in the construction of power plants and other structures that experience operating temperatures in excess of 500 degrees C. The commonly used universal slopes approach for fatigue life modeling for which the strain life computation employs the standard-Coffin Manson relationship is compared to a statistical methodology using a distribution function frequently used in structural reliability. The proposed distribution function for characterizing the fatigue life is a generalized Weibull distribution function that empirically incorporates load history and damage accumulation.