Fatigue lifetime prediction and reliability assessment for fusion blankets under pulsed operational loading

Abstract The breeding blanket is a critical component of magnetic confinement fusion reactors. The reliability, availability, maintainability, and inspectability (RAMI) of the blanket has been recognized as a major challenges on the pathway to commercial fusion demonstration reactors (DEMO). However, the absence of blanket failure modes and failure rate data under fusion environment has significantly impeded progress in blanket RAMI research. To address this critical knowledge gap, the present investigation focuses on the dominant failure mechanism induced by pulsed operational loading - specifically fatigue failure, and proposes a novel probabilistic framework to predict lifetime and assess reliability for fusion blankets. The framework incorporates three key elements: (1) structural material damage prediction model grounded in the degradation mechanisms, (2) operational loading spectra characterization, and (3) blanket lifetime prediction and reliability assessment. The framework enables probabilistic evaluation of fatigue lifetime while generating essential reliability metrics, including failure rates and mean time between failures (MTBF). Validated through application to the Water-Cooled Ceramic Breeder Blanket (WCCB) design for the Chinese Fusion Engineering Testing Reactor (CFETR), the proposed framework demonstrates the potential to overcome existing data limitations in blanket RAMI research, offering substantial support for the blanket design optimization.