Specimen miniaturization for fracture toughness determination of RPV materials - a combined experimental / modeling investigation

In the context of the structural integrity assessment of nuclear reactor pressure vessels (RPV), the mini-CT specimen is gaining worldwide attention as one of the ideal miniaturized geometries for measuring fracture toughness. This geometry is particularly appealing for several reasons. Firstly, up to eight mini-CT specimens can be machined out of one broken Charpy specimen, thus avoiding the need to consume additional surveillance materials. Secondly, the geometry permits specimen reorientation, which facilitates the investigation of anisotropic effects. Another considerable advantage that enables this geometry to replace large ones is that the mini-CT geometry has shown potential to produce fracture toughness results equivalent to large samples. However, the miniaturization of test geometries induces a loss of constraint, which may lead to deviation of fracture toughness values when compared to those measured from larger geometries. In addition, due to the reduction in size, this geometry does not fully satisfy the ASTM requirements with respect to the varies ratios of the specimen dimensions. In particular, the requirements relative to the pre-crack front curvature and initial crack length are challenging, leading to a large number of mini-CT specimens not fulfilling the standard requirements and therefore being considered invalid. Therefore, in this thesis, three main issues in using mini-CT geometries: 1) effect of constraint loss, 2) effect of pre-crack front non-uniformity, and 3) effect of initial crack length are investigated to promote the standardization of the mini-CT geometry in both brittle and ductile regimes. Experimental investigations and finite element simulations are used as the two main analytical methods. In addition, micromechanics-based models are also introduced: the Rice-Tracey model and the Thomason model are used to analyze ductile fracture, and the Particle distribution model is used to analyze brittle fracture. The results show that the effect of constraint loss on fracture toughness can be compensated by applying an appropriate scaling factor in the ductile fracture regime and a new size correction in the brittle fracture regime. The effect of pre-crack non-uniformity can be considered limited, indicating that the requirements for pre-crack front curvature could be relaxed and therefore avoid unnecessary rejection of meaningful data from mini-CT specimens. The requirements for the initial crack length in mini-CT geometry should remain consistent with the ASTM standard. However, in the brittle fracture regime, it may be necessary to apply additional constraint corrections to the fracture toughness measured from mini-CT specimens with a0/W below 0.5 in order to achieve a higher level of accuracy in T0 values. The results demonstrate that, by implementing the adjusted specifications, the mini-CT geometry is capable of producing reliable and accurate fracture toughness measurements.