Abstract: According to the essential hypothesis and principle of probabilistic fracture mechanism, the advanced method of estimating generic parameters of EIFS distribution is presented. The new method calculates TTCI values at different reference-crack-sizes with growth curve of small crack and estimates TTCI distribution parameters with maximum likelihood estimation. Then, the generic upper limit of EIFS distribution and crack growth parameter under different stress levels are estimated with constrained least squares fitness. Based on these, the generic parameters of EIFS distribution are determined according to generality condition of EIFS distribution parameters. Compared to the old method, the new one makes adequately use of the information of crack growth and distribution function, which improves the rationality and reliability of estimated TTCI and EIFS distribution parameters. The IFQ analysis of a structure detail proves that the advanced method is feasible and convenient for application.

Key words: Initial Fatigue Quality (IFQ), Equivalent Initial Flaw Size (EIFS), Maximum Likelihood Estimation (MLE), Time to Crack Initial (TTCI)