On the history and status of reactor pressure vessel steel ductile to brittle transition temperature shift prediction models

Abstract This paper melds an overview of the history of Reactor Pressure Vessel (RPV) embrittlement research, focusing on predicting ductile-to-brittle transition temperature shifts (ΔT), along with an assessment of the current status of these efforts, especially for extended life operation. The 60-year history of RPV research reveals remarkable progress on a very complex and challenging problem that has, for several decades, been a paradigm for a ‘science in service of engineering’ approach to a critical technological challenge. This research has laid the foundation for properly analyzing modestly accelerated materials test reactor (MTR) data to make robust ΔT predictions beyond the current low flux (φ) power reactor surveillance database. We show that most current models, that are accurate at lower fluence (φt), systematically and significantly underpredict ΔT at high φt, largely owing to the currently unaccounted for contribution of late blooming MnNiSi precipitates (MNSPs). We propose a simple empirical approach to predicting ΔT between φt ≈ 4 × 1023 n/m2 (the currently reliable ΔT model limit) and 14 × 1023 n/m2 (for extended life). The method is shown to be empirically robust, and is supported by a microstructurally informed physical model. In addition to quantifying the role of MNSPs, important observations include approximately linear ΔT dependence at high φt (versus the previous ≈ √φt trend at lower φt), and a diminution of the effect of φ. The decreased φ effect at high φt has very important implications for the use of accelerated MTR data to predict service relevant ΔT.