The probabilistic modeling of fatigue crack growth is reviewed, primarily from 1980 to the present. Attention is directed to evolutionary models that start with the randomization of a deterministic differential equation by the introduction of random variables or random functions, and to models that directly generate the evolution of distributions. As a result of the statistical analysis of data, reasons are advanced in favor of increasing the dimension of the observables employed.

This content is only available via PDF.
You do not currently have access to this content.