Geometry and Material Property Uncertainty Model for Fatigue Life Predictions

The objective of the present work is the development and demonstration of a statistical model for the prediction of fatigue life of metal and alloys under different probability and confidence levels. This model is based on the assumption that fatigue life of a component follows log normal distribution. It describes an analytical model which is derived from energy theorem and coupled with the uncertainty associated with material properties and geometrical parameters to estimate fatigue life and crack growth at different level of probability and confidence level. A substantial amount of published data has been used to validate the proposed model for predicting S-N curve as well as fatigue crack growth of steel, aluminum alloys, copper alloys and titanium alloys