Fisheye failure analysis and life design approach for case-carburized gear steel based on statistical evaluation of defect size

The high cycle fatigue (HCF) and very high cycle fatigue (VHCF) properties of a case-carburized gear steel were investigated under tension and compression, and the distribution characteristics of defect size was analyzed using four statistical functions including Statistics of Extreme Values (SEV), Generalized Pareto distribution (GP), Generalized Extreme Values (GEV) and Exponential Generalized Pareto distribution (EXPGP). Results show that the interior inclusion-induced failure with the existence of fisheye becomes the predominant failure mode in HCF and VHCF regimes. With the increasing of applied stress amplitude, the effect of crack size within the fisheye on fatigue life is gradually reduced. Based on this, a fatigue strength prediction model associated with the variation of crack size was developed. At a given probability, the predicted fatigue strength is the highest by the estimated maximum inclusion size using the GEV distribution, then followed by using Gumbel distribution and by using GP distribution, and the lowest by using EXPGP distribution, which is inversely proportional to the evaluated maximum inclusion sizes. Based on the good agreement between the predicted and experimental results, the proposed approach based on statistical evaluation of defect size can be well used to predict the HCF and VHCF fatigue lives with interior fisheye failure.