Abstract: The analysis of fatigue crack growth damage accumulation is one of the most important tasks to the durability and damage tolerance design and the fatigue reliability design of structure. It is widely recognized that fatigue crack growth is fundamentally a random phenomenon, which can be predicted only in terms of probability. The phenomenon is much more complex if a structural element is subjected to random loading because of the fatigue crack growth retarding and accelerating effects. In thsi paper, by using the theorem of probabilistic fracture mechanics and stochastic process, a statistical model coupled with the root mean square method has been developed for the predicting of fatigue crack growth damage accumulation under random loading. Both the distribution of crack length at any given life and the distribution of load cycles to reach any given crack length are predicted. A special experiment, in which specimen size is 96 and specimen material is LY12CZ, an aluminium alloy, is conducted under a fighter random loading for the purpose of verification. A good correlation is obtained between the predicted distribution and the experimental results.

Key words: probabilistic fracture mechanics, stochastic process random loading, fatigue crack growth