A TWO CRITERIA STATISTICAL MODEL FOR TRANSITION FRACTURE TOUGHNESS

Abstract

Abstract— A two criteria model is proposed to explain the nature of the fracture toughness scatter in the transition regime for ferritic steels. In the early part of the transition the model describes the data by a normal statistical distribution. Here the mean of the distribution is constant with specimen size and the variance of the distribution varies inversely with test specimen size. At a higher temperature, in the middle of the transition, the model results in a Weibull distribution in which both the mean and variance of the distribution vary inversely with change in specimen size. For both of these cases the fracture toughness scatterband from one specimen size can be used to predict the scatterband for another specimen size provided that the thickness constraint is not changed. Between these two regions the model results in a mix of normal and Weibull distributions for which the distribution from one size cannot presently be used to relate to another specimen size. The model is based upon the distribution of the transition fracture toughness data not on the observed microstructural behaviour; however, it is consistent with microstruc‐ tural observations. The model does not completely describe all observed data trends; however, it appears to describe observed behaviour better than single criterion models.