High-Performance Manufacturing of Canned Main Coolant Pump—Part II: Reliability Evaluation

Abstract Major barriers that hinder application of advanced manufacturing technologies are unknown reliability resulted by the high-performance manufacturing (HPM) of a canned main coolant pump (MCP) toward desired functional performance in nuclear power plant (NPP). This work provides an expanded HPM framework to the integrated design and processing of the canned MCP taking account of system-components crosslinking on material-product-process linking via a pivotal role of experimentally characterized surface integrity. Hereby, Part I presents synergistical design and processing optimization of the canned MCP prototype manufactured by additive manufacturing, near net shape forging, joining/cladding, and surface modification/coating. In Part II, the reliability of canned MCP prototype manufactured by the optimized advanced processes is evaluated by means of a quality control method in the functional performance inequality under entire service conditions with 60-year lifespan. The respective service conditions of key components range from the empirically allocated entire service conditions to the quantitatively simulated ones in the coupled design and processing modeling for the extreme dynamic performance degradation of canned MCP in the service by the experimentally characterized surface integrity. The critical failure criteria of key components depend on the varied respective service conditions, to recognize the major failure modes including wear, corrosion, impact, fatigue, and fracture. An innovative probability-service-life map of the key components covers the major failure modes from the failure probabilistic density functions following the experimental relations of specific performance with the surface integrity change in lognormal and/or Weibull distribution under the varied specific service conditions to the critical failure criteria. The varied respective service conditions are optimized according as a dynamic performance residual simulated in the coupled modeling during the service cycles to determine the normalized probabilistic service lives of aged key components in the probability-service-life maps. The linear cumulative damage by the normalized probabilistic service lives is updated to judge the failure of each aged key component in service. A failure probabilistic density function of the canned MCP system establishes by the linear cumulative damage of all the aged key components in a unified operation time scaling of service years. Finally, the degree of reliability of canned MCP prototype manufactured by the optimized design and processing parameters is calculated by the failure probabilistic density function of the system for the 60-year service requirement.