Acoustic emission-based fatigue crack growth prediction

Abstract

Acoustic Emission (AE) is a non-destructive testing (NDT) with potential applications for locating and monitoring fatigue cracks during structural health management and prognosis. To do this, a correlation between acoustic emission signal characteristics and crack growth behavior should be established. In this paper, a probabilistic model of fatigue crack length distribution based on acoustic emission is validated. Using the results from AE-based fatigue experiments the relationship between AE count rates and crack growth rates is reviewed for the effect of loading ratio. Predictions of crack growth rates based on AE count rates model show reasonable agreement with the actual crack growth rates from the test results. Bayesian regression analysis is performed to estimate the marginal distribution of the unknown parameters in the model. The results of the Bayesian regression analysis shows that the results are consistent with respect to changes in loading ratio. Additional experimental evidence might be required to further reduce the uncertainties of the proposed model and further study the effects of the changes in loading frequency, sample geometry and material.