Statistical models for prediction of the fatigue crack growth in aircraft service

One of the most important problems in the fatigue analysis and design of aircraft structures is the prediction of the fatigue crack growth in service. Available in-service inspection data for various types of aircraft indicate that the fatigue crack damage accumulation in service involves considerable statistical variability. The statistical nature of the fatigue crack growth is attributed to, among others, two most important factors: (i) the statistical nature of service loads and environments experienced by aircraft structures and (ii) the inherent fatigue crack growth variability of materials. The objectives of this paper are to (i) describe possible statistical models to deal with the crack growth variability, (ii) point out their applications, (iii) suggest an invariant embedding technique for statistical decision making on the basis of fatigue crack growth models and (iv) propose an approach to evaluating the goodness of these models.