Analysis and verification of fatigue reliability variation under two-stage loading conditions

The effective life of a specimen is calculated as the time from when the specimen is in its operating conditions to the time of its failure and reliability indicates probability of survival of the specimen or product. In material science, this effective life is represented as the number of loading cycles until the failure of the specimen. Life of any product or specimen is inversely proportional to the load (stress or strain) applied on it. Practically, external loading is mostly fluctuating and various levels of loads are applied on the same specimen causing the specimen to fail over time and this is referred to as fatigue failure under variable amplitude loading. Researches in the past have even shown that life prediction under variable amplitude loading cannot be fit to any conventional method. Besides the constraints faced by modeling under variable amplitude loading, it is important to derive a method which can predict and analyze reliability variations under various conditions.;In this study we have combined reliability prediction methods from past researches and presented a method to predict the fatigue reliability of a specimen under two-stage loading conditions with the help of failure data under constant amplitude loading, two dimensional probabilistic Miner's rule and Weibull analysis. Corresponding reliability values of the specimen at different stages are calculated. A significant relationship between reliability of the specimen and the change in stress level is derived with the help of test data results from simulation. The consistency of the obtained reliability values is examined by further calculating the Miner's verification coefficient and the variation in consistency is also studied.