Experiments and stochastic model for propagation lifetime of railway axles

Inspection intervals of critical safety components should be calculated on the basis of sophisticated crack growth algorithms able to describe crack propagation under service loads. The present paper deals with a probabilistic application of the NASGRO crack growth algorithm to estimate the propagation lifetime of railway axles. The analysis of experimental crack growth and threshold data of an A1N steel has enabled us to propose two random variable models for the estimation of propagation lifetimes, discussing their statistical properties and their implications for lifetime estimates.