Intrinsic statistical characteristics of fatigue crack growth rate

Abstract

An intrinsic probabilistic description of the whole range fatigue crack growth rate is proposed by accounting For the interaction of the quasi-static crack growth, the crack closure, and the crack growth threshold. The crack growth rate is shown to be reasonably well represented as a function of effective K factor when the quasi-static crack growth is empirically accounted for. It was found from 2024-T3 aluminium alloy experimental crack growth rate data that the scatter of crack growth rate can be reasonably well represented as a Gaussian distribution with nearly constant standard deviation in the normal direction of the mean crack growth curve in log-log coordinates when the crack growth rate is related to the effective K factor. A system of mathematical formations is derived to transform the distribution of the experimental data in the normal direction of the mean crack growth curve to either the crack growth rate or the effective K factor. Significant improvement has been achieved in solving the scatter of crack growth rate as a function of the effective K factor based on the proposed method. The mechanical effect has been minimised in this model so that the proposed relation reasonably represents the material dependent resistance to the fatigue crack growth driving force. This material dependent relation provides a good base-line in probabilistic analyses of fatigue crack growth using various stochastic process methods, i.e. the Monte-Carlo crack growth simulation, the Markov and Markov chain crack growth simulation, or as a starting point for other statistic fatigue life analyses.