Tham khảo tài liệu 'metal fatigue part 13', kỹ thuật - công nghệ, cơ khí - chế tạo máy phục vụ nhu cầu học tập, nghiên cứu và làm việc hiệu quả | 344 Appendix A Table The estimated value of v re max and the lower bound of fatigue strength TW Number of specimens 1 10 100 ơw MPa 475 464 456 Voreamax 100 a 100 b X im f Substituting each value of y areamm into Eq. the lower bounds of the fatigue strengths ƠW as a function of the Vickers hardness Hy 500 are predicted as follows Owl 7v 120 Vãrêãmax 6 Owid 7v 12O 6 475 MPa 7W 1O 7v 120 6 464 MPa Owi ioo Hv 120 l 6 456 MPa The units are ƠW1 MPa Hy kgf mm2 ựõrẽãmax im. Table shows the values of Jarea and ơw . V max m Appendix A 345 A8 Optimisation of Extreme Value Inclusion Rating EVỈR From the results shown in the previous sections one could draw the wrong conclusion that the estimation of the maximum inclusion or defect in a component is a very simple procedure which can be based on a few measurements on small areas. The key point is to optimise the sampling procedure in order to obtain significant and reliable estimates of extreme defects. The first problem that has to be addressed is the uncertainty in the x T estimation. The value of x T can be easily calculated by Eq. A from the parameters À and 8 of the actual sample. However these values are only estimates of the true parameters of the population of defects from which the sample was taken. The parameter estimates are more precise with the increasing number n of maximum defect examined and they tend to true values when n oo. It follows that the number of defects to be collected with extreme value sampling has to be chosen in order to make an x T estimate precise enough for fatigue strength calculations. This can be done by obtaining more precise estimates of x T with the maximum likelihood method and by choosing the minimum number of defects with the map shown in Fig. 3 1 Shape ratio Ỗ Ằ Figure Maps for optimising the precision of x T estimates for maximum likelihood statistical analysis 3
