A Probabilistic Model for Describing Short Fatigue Crack Growth Behavior of LZ50 Steel

Fatigue damage process of metal components and structures with smooth surface belongs primarily to the short fatigue crack stage. To characterize the random growth behavior of short fatigue crack and to apply the crack growth rate model for engineering safety assessment, a probabilistic model is proposed with consideration of the test data scattering regularity. This probabilistic model is based on the multi-microstructural barriers' model and can describe the deceleration behavior of crack growth rate during the total short fatigue crack propagation period. To take the statistical characteristics of the test data into account, the idea from the general maximum likelihood method which is widely used in parameter estimation of fatigue S - N curves and epsilon - N curves is inherited. While estimating the parameters of the probabilistic model, the conventional correlation coefficient optimization method is extended for calculating the parameters of both the mean and the standard deviation curves. Analysis of the test data of LZ50 steel indicates the reasonability and applicability of the proposed model.