13th International Conference on Fracture June 16–21, 2013, Beijing, China -9- deformation.flow accompanying an increasing crack tip opening displacement. During fatigue cycles increase not only the crack size became largebut also the peak values of corresponding RDF curves reduced continuously. Figure 12. The RDF curves at 300 K, 900 K and 1100 K t /ns 0 2 4 6 8 10 12 14 16 RDFmax 4 6 8 10 12 14 16 200 K 300 K 500 K 700 K 900 K 1100 K Figure 13. Relationship between the ultimate values of RDF and time Figure 14 shows the RDF curves via temperature change at 5, 10 and 15 ns respectively. The ultimate values of RDF at different temperatures are plotted in Figure 15. Based on thermal activation theory and simulation results, the effect of temperature on crack growth can be taken into account by an expression in an exponential function form similar to expressions (1) and (2) which describes crystal plasticity. The coefficent and index values in expression are determined by the RDF peak value obtained from atomistic simulations. Here are some data processing results. For time 5 ns: T m e C T 99.53 ( ) 5.2191 − = , with correlation coefficient R=0.91. For time 10 ns: T m e C T 234.89 ( ) 4.6203 − = , with correlation coefficient R=0.99. For time 15 ns: T m e C T 225.61 ( ) 4.9712 − = , with correlation coefficient R=0.97. High regression correlation coefficients imply that the explanation and the used function form for considering the temperature influence on crack growth are appropriate. Figure 14. RDF curves via temperature change at 5 ns, 10ns and 15 ns Figure 15. Relationship between the ultimate values of RDF and temperatures
RkJQdWJsaXNoZXIy MjM0NDE=