ICF10B

The results here are also compared with some analytical and experimental approximations available [24-28]. The predictions in the figures were made by assuming that the total maximum SIF is the superposition of the maximum applied SIF and the residual SIF. The experimental results obtained here fall in between the Rice and Irwin analytical solutions [26,27]. The experimental results are also shown compared to calculations based on constants experimentally obtained by Birol [25], which was based on an etching method, and Bathias [28], which was based on a microhardness method. SUMMARY AND CONCLUSIONS Accurate predictions of fatigue crack growth rates through residual stress fields were obtained from the superposition of residual stresses and applied stresses using the weight function approach. Both the residual stresses and the applied stresses were converted to stress intensity factors independently and then combined using the basic principle of linear elastic superposition. The specimens with initial tensile residual stresses show acceleration of the crack propagation rates, which compare well with predictions based on models. Evaluation of the results lead to the following conclusions: 1. The residual stresses in a specimen can be used, with the weight function, to compute a residual stress intensity factor that a crack will experience as it propagates through the residual stresses. Then, with a crack propagation law, one can predict the fatigue crack growth rate of a crack growing into a tensile residual stress field. Superposition may not work satisfactorily if either the residual or applied stress intensity factors are compressive. If the net stress intensity factor is decreasing as the crack grows, the plastic zone size is decreasing which can cause the crack to behave differently than the growth model. 2. Tensile residual stresses accelerate a fatigue crack by increasing the load ratio. This has two effects. First, materials display a load ratio sensitivity, usually quantified by the Walker exponent γ, such that cracks grow faster at a higher load ratio. Secondly, at a higher load ratio cracks will transition into the more rapid Stage III growth regime at a lower stress intensity range. This effect can be seen in Figure 5 as the point where the TC model deviates from the Walker model predicting higher crack growth rates and better representing the experimental data. 3. The three-component model of Saxena gives a good prediction of fatigue crack propagation rates when used with the superposition of the applied and residual stress intensity factors. It captures both of the load ratio sensitivity and the transition to Stage III growth and hence makes a better prediction for the crack growth rate than the Walker model. 4. The size of the plastic zone can be measured with a small synchrotron x-ray beam along a fatigue crack by measuring the diffraction line broadening of the material perpendicular to the crack faces. The superposition of the residual stress intensity factor and the maximum applied stress intensity factor during fatigue crack growth correctly predicts the shape of the plastic zone size variation as a fatigue crack grows into a tensile residual stress field. The measured size of the plastic zone falls between the theoretical predictions of Rice and Irwin, which give upper and lower bounds, respectively. ACKNOWLEDGEMENTS The authors would like to acknowledge the University of Missouri Research Board for funding this research. They also express thanks to J.D. Almer and J.B. Cohen for performing the initial residual stress measurement at Northwestern University. The assistance of D.R. Haeffner and P.L. Lee with the synchrotron x-ray measurements is gratefully acknowledged. The use of the

RkJQdWJsaXNoZXIy MjM0NDE=