Application of the Master Curve Approach for Abnormal Material Conditions

Abstract

The range of applicability of Master Curve testing Standard ASTM E 1921 is limited to macroscopically homogeneous steels with “uniform tensile and toughness properties”. A majority of structural steels appear to satisfy this requirement by exhibiting fracture toughness data which comply with the assumed KJc vs. temperature dependence and scatter within the specified validity area. As indicated in ASTM E 1921 a criterion for material macroscopic inhomogeneity is often applied using the 2% lower bound (possibly also the 98% upper bound). Data falling below this 2% lower-limit curve may be an indication of material inhomogeneity or susceptibility to grain boundary fracture. When this situation occurs, it is recommended to analyze the material with the so-called SINTAP procedure, which is intended for randomly inhomogeneous materials to assure a conservative lower-bound estimate. When a data set distinctly consists of two or more different data populations instead of one (due to variation of irradiation dose or specimen extraction depth, for instance) adoption of a bimodal (or a multimodal) Master Curve model is generally appropriate. These modal models provide information if the deviation of distributions is statistically significant or if different distributions truly exist for values of reference transition temperature, T0, characteristic of separate data populations. In the case of data sets representing thick-walled structures (i.e., reactor pressure vessels), indications of abnormal fracture toughness data can be encountered such that material inhomogeneity or fracture modes other than pure cleavage should be suspected. A state-of-the-art review for extended, non-standard Master Curve data and techniques highlights limits of applicability in situations where the basic ASTM E 1921 procedure is not appropriate for material homogeneity or different fracture modes.