A NOVEL METHOD FOR FRACTURE TOUGHNESS ASSESSMENT OF INHOMOGENEOUS FERRITIC STEEL WELDMENTS USING BIMODAL MASTER CURVE ANALYSIS P. Nevasmaa, A. Laukkanen, T. Planman and K. Wallin VTT Industrial Systems, P.O. Box 1704, FI-02044 VTT, Finland ABSTRACT The basic Master Curve (MC) method for analysis of brittle fracture test results is intended only for macroscopically homogeneous ferritic steel. In reality, steels and welds often contain inhomogeneities that distort the standard MC analysis. The structural integrity assessment procedure SINTAP contains a lower tail modification of the MC analysis, enabling conservative lower bound fracture toughness estimates to be determined also for inhomogeneous material. Such estimates, however, only describe the fracture toughness of the more brittle constituent. The deficiency of SINTAP, in this respect, lies in its inability to provide any information from the more ductile material. Therefore, a probabilistic description of the complete material is not possible. This paper introduces a new extension of the MC analysis for inhomogeneous material: a bimodal MC analysis method that describes the fracture toughness distribution as the combination of two separate MC distributions. This bimodal distribution model is shown to describe successfully the weld heat- affected zone (HAZ) fracture toughness data sets that generally exhibit substantial microstructural inhomogeneity. This is especially the case with multipass weldments containing local brittle zones (LBZ). 1 INTRODUCTION The basic Master Curve (MC) method for analysis of brittle fracture test results as defined in ASTM E1921-02 is intended for macroscopically homogeneous ferritic steels only. In reality, the steels in question are seldom macroscopically fully homogeneous. The steels fracture toughness may depend on the specimen location in the sample. For example, thick plates and forgings may have very different fracture toughness at plate center and close to surface. The inhomogeneity may be deterministic or random (or a mixture of both) in nature. Deterministic inhomogeneity can be accounted for, provided that the specimen extraction histories are known and enough specimens are tested. Random inhomogeneity is much more difficult to handle. The structural integrity assessment procedure SINTAP contains a lower tail modification of the MC analysis. This enables conservative lower bound type fracture toughness estimates also for inhomogeneous materials. The problem is that the SINTAP method, does not provide information of the tougher material. Therefore, a probabilistic description of the complete material is not possible. Here, a new comparatively simple extension of the MC is introduced for inhomogeneities governed by two separate MC distributions. The extension is shown to be well suited to describing weld heat-affected zone (HAZ) data. 2 THE MASTER CURVE METHOD The Master Curve (MC) method incorporates descriptions for (i) cumulative failure probability distribution of brittle cleavage fracture in a macroscopically homogeneous ferritic steel, (ii)