The problematic nature of the application of stochastic crack growth models in engineering design

Two major families of solution methods for the stochastic crack growth rate under repeated loading are reviewed. These solutions are required in order to calculate the probability of failure of cracked structural elements. It is shown that when the parameters in one of these families are randomized, the dimensions of some variables become random, an outcome which has no physical meaning. Suggestions are made as to how these models should be expressed so that the physical interpretation of the solution process remains meaningful.