Abstract
Current guidance considers that uniaxially loaded specimen with a deep crack is used for the determination of the ductile-to-brittle transition temperature. However, reactor pressure vessel is under biaxial loading in real and the existence of deep crack is not probable through periodic in-service-inspection. The elastic stress intensity factor and the elastic-plastic J-integral which were used for crack-tip stress field and fracture mechanics assessment parameters. The difference of the loading condition and crack geometry can significantly influence on these parameters. Thus, a constraint effect caused by differences between standard specimens and a real structure can over/underestimate the fracture toughness, and it affects the results of the structural integrity assessment, consequentially. The present paper investigates the constraint effects by evaluating the master curve $T_0$ reference temperature of PCVN (Pre-cracked Charpy V-Notch) and small scale cruciform specimens which was designed to simulate biaxial loading condition with shallow crack through the fracture toughness tests and 3-dimensional elastic-plastic finite element analyses. Based on the finite element analysis results, the fracture toughness values of a small scale cruciform specimen were estimated, and the geometry-dependent factors of the cruciform specimen considered in the present study were determined. Finally, the transferability of the test results of these specimens was discussed.