Applicability of the modified Ritchie–Knott–Rice failure criterion to transfer fracture toughness Jc of reactor pressure vessel steel using specimens of different thicknesses—Possibility of deterministic approach to transfer the minimum Jc for specified specimen thicknesses

Abstract

The validity of the modified Ritchie–Knott–Rice failure criterion to transfer the fracture toughness Jc of a reactor pressure vessel steel material was examined using specimens with thicknesses in the range of 8–254mm. Here, the modified Ritchie–Knott–Rice criterion is the (4δt, σ22c) criterion, which predicts the onset of cleavage fracture when the crack-opening stress σ22 measured at a distance from the crack tip equal to four times the crack-tip opening displacement δt, denoted as σ22d, exceeds a critical value σ22c. Large-strain elastic–plastic finite element analysis was performed to reproduce the fracture toughness (Jc) test results of 21 SE(B) specimen types and 141 specimens satisfying the ASTM E1921 requirements of M=(b0σYS)/Jc⩾30 and 0.45⩽a/W⩽0.55, where b0 is the initial ligament size, σYS is the yield stress, and a/W is the crack depth-to-width ratio. The material was reactor pressure vessel ASTM A533 Grade B class 1 steel, and tests were conducted in the ductile-to-brittle transition temperature region. The results show that the modified Ritchie–Knott–Rice failure criterion is valid for specimens of various thicknesses, although in this study, the constraint differences were small. The critical value σ22c showed only 4.6% variation, and its probabilistic nature appeared to be negligible. The effect of the test specimen thickness on Jc was concluded to result from the inability of J to characterize the crack-tip stress fields accurately. Although further investigation of the application limit of the (4δt, σ22c) criterion is necessary, the results appear to show that the critical value σ22c is a deterministic value corresponding to the minimum Jc observed for a specific specimen, which can be determined by the specimen geometry and tensile property of the material. Because application of the (4δt, σ22c) failure criterion requires only a single specimen or set of test data, whereas calibration of the Weibull stress in the probabilistic approach requires many specimens, the deterministic approach using the (4δt, σ22c) criterion may be suitable for use in place of the probabilistic approach using the Weibull stress in certain circumstances, such as in the cases that the minimum Jc is of main interest.