Abstract
This paper is the second in a series of two papers that describe benchmark comparisons of 14 probabilistic fracture mechanics codes. The benchmark problems focus on determining the leak-before-break behavior of a nickel-based alloy weld in a large-bore piping system of a pressurized water reactor. The modeled degradation mechanism is primary water stress corrosion cracking. In Part 1, which is covered in the technical paper PVP2022-84724, the benchmark problem was deterministically analyzed with the probabilistic fracture mechanics codes using their models for crack growth rates, stress-intensity factors, crack-opening displacements, crack transition from inside surface-breaking cracks to through-wall cracks, leak rates, and crack stability. Several output quantities of interest relevant to leak-before-break behavior were compared. In Part 2, the benchmark problem was probabilistically analyzed with the probabilistic fracture mechanics codes by introducing uncertainties in key input parameters, such as the crack growth rate and welding residual stresses. The probabilities of leak and rupture as a function of time were then compared accounting for the effects of leak detection and inservice inspection. Interpretations of the probabilistic comparisons are provided to assess scatter among the probabilistic fracture mechanics code predictions. These interpretations are supported by an enhanced understanding of the different modeling approaches as gleaned from the deterministic benchmark comparisons.
