Fatigue Model for Composites Based on the Cycle-by-cycle Probability of Failure: Implications and Applications

Abstract

Further implications of a fatigue model based on the cycle-by-cycle probability of failure are presented. A model for the residual strength is developed based on a linear relation between the residual strength at any cycle level and its derivative. This model is shown to be in excellent agreement with published test results. The model is also shown to lead to a constant cycle-by-cycle probability of failure under constant amplitude loading thus verifying the assumption of constant probability of failure made previously. The model is then used to construct constant life (Goodman) diagrams for composite structures and its predictions are compared to published test results. The model follows the test data well but needs further improvement for negative R values. The approach is then extended to the determination of the truncation level required for a structure to meet a certain fatigue life. The truncation level is a function of R ratio and the amount of statistical scatter for static tension and compression tests. Using representative values for the scatter, the predicted truncation level of 38% is shown to cover most applications. Possible reasons for discrepancies and areas where more work is needed are identified.