The statistical analysis of fatigue data using the Weibull distribution
Abstract:ESDU 21004 presents methods of using statistical analyses to interpret fatigue data and to predict the life of a component subject to fatigue. In this Data Item, an introduction to continuous distributions is given including an overview of probability density functions, cumulative distribution functions, the reliability function and the hazard rate function. A brief overview to the normal distribution and the log normal distribution is also provided for reference. The statistical analysis of fatigue life data using the Weibull distribution is considered in detail in this Data Item. An examination of reliability theory, the influence of the Weibull parameters on the distribution, the various sources of uncertainty and the use of confidence bounds when interpreting the data are given. The Weibull weakest-link probability theory and its use for analysing the size effect is discussed. The introduction to analytical and computational methods that are used to plot, examine and compare fatigue data using the Weibull distribution are outlined and two fully worked examples are provided. The analytical and computational methods that are used to plot, examine and compare fatigue data using the Weibull distribution are outlined, and two fully worked examples are provided.
|Data Item ESDU 21004|