Cast Austenitic Stainless Steel (CASS) can be subject to thermal aging when exposed to operating temperatures of PWR (Pressurized Water Reactor). During this process, the yield and tensile strengths can increase slightly, but its fracture toughness can decrease substantially. A simple prediction model to predict the thermal aging effect on the maximum load of a cracked CASS pipe was previously proposed by the authors. In this study, the simplified prediction model is validated against experiment results for cracked CASS pipes (from the JNES report). The results show that the model accurately predicts the experimental maximum load tendency, provided that the strain-hardening exponent (n) of the unaged material should be appropriately determined. Since the value of n can vary depending on the fitting range, a sensitivity analysis of the material parameters is needed for the practical and reliable application of the prediction model.

This study experimentally investigates the surface residual stress characteristics of helical steam generator tubes subjected to cold bending for small modular reactor (SMR) applications. To simulate the manufacturing process, Alloy 690 tube specimens were fabricated using a free-form cold bending machine capable of continuously varying curvature. The bending condition was parameterized using the dimensionless bending radius ratio (R/D) and the dimensionless pitch parameter (P/D). Surface residual stresses were measured non-destructively by X-ray diffraction using the single incident angle exposure (cosΨ) method. Residual stresses were evaluated at four circumferential positions (0°, 90°, 180°, and 270°), corresponding to flank, intrados, and extrados regions of the bent tube. The results show that compressive residual stresses are dominant at the flank and intrados, while tensile residual stresses are observed at the extrados. The normalized extrados tensile residual stress shows a decreasing tendency with increasing R/D, while the effect of P/D varies depending on the range, with a partial reduction observed in certain conditions. These findings provide fundamental data to support design and manufacturing guidelines aimed at minimizing tensile surface residual stress and enhancing SCC resistance in SMR steam generator applications.

Stress Corrosion Cracking (SCC) is a critical degradation mechanism for austenitic stainless steels in Light Water Reactor (LWR) environments. While machine learning shows promise in predicting Crack Growth Rates (CGR), its 'black-box' nature makes it difficult to elucidate underlying causal relationships. To address this, the present study developed an integrated CatBoost-based model with Shapley Additive exPlanations (SHAP) value estimation as a powerful analytical tool for thermodynamic interpretation. By projecting estimated results onto E-pH diagrams and dynamic thermodynamic driving force (∆ECP) maps, this Explainable Artificial Intelligence (XAI) approach was shown to successfully identify hidden thermodynamic patterns otherwise difficult to detect in complex empirical databases. By clearly distinguishing the magnetite-hematite phase boundaries governing CGR acceleration and retardation, this study highlights XAI as an effective tool for uncovering physical principles in nuclear materials research.

In 1989, NRC GL 89-13 (Service Water System Problems Affecting Safety-Related Equipment) to verify the cooling performance of the service water system of nuclear power plants was the starting point for the reliability regulations of nuclear power plant cooling water systems. To address this issue, KHNP (Korea Hydro & Nuclear Power) developed "Heat Exchanger Integrated Management Program" in 2010, using which the performance and manage performance data of service water heat exchangers can be analyzed based on the EPRI "Service Water Heat Exchagner Testing Guidelines (1998)". KHNP launched another research program in 2010 to improve the reliability of heat exchanger performance evaluation results and to advance the algorithm. This paper summarizes the outcomes: (1) confirm the reliability of the performance evaluation results, (2) overcome the limitation of evaluating performance under low heat load conditions, and (3) make it possible to evaluate major heat exchangers that were not previously managed by adding a plate heat exchanger performance model.

This paper investigates the effect of cycle counting methods on elastic-plastic fatigue evaluation based on JSME Code Case NC-CC-008 under seismic random loading. Dynamic cyclic elbow test and elastic-plastic finite element analysis were performed for SA234 WPB carbon steel and SA403 WP316 stainless steel elbows. The hoop strain history at the critical location was extracted from elastic-plastic FE analysis results, and three different cycle counting methods (simple-range counting, peak-valley counting, and rainflow counting) were applied. For both materials, simple-range counting gave the least conservative results, and peak-valley counting gave the most conservative results, although evaluation results were not so different. For SA234 WPB, the calculated cumulative usage factor (CUF) was 21.8, 25.1, and 27.8 for simple-range counting, rainflow counting, and peak-valley counting, respectively, while the corresponding values for SA403 WP316 were 13.8, 15.0, and 15.2. Based on these results, simple-range counting can be recommended for elastic-plastic analysis-based assessment due to its simplicity and easy applicability.

This study aims to develop a quantitative depth sizing technique for circumferential cracks that may occur in feedwater heater tubes of nuclear power plants using bobbin-type eddy current testing (ECT). A total of 84 specimens, consisting of 52 artificial flaw specimens and 32 natural crack specimens, were fabricated and obtained for this purpose. Each specimen was evaluated using 18 bobbin ECT measurement techniques derived from combinations of inspection frequency (280, 140, 70 kHz), calibration standard type (ASME flat-bottom hole and circumferential EDM notch), and phase angle measurement method (Vpp, MxR, Vmx). Regression analyses were performed between the actual flaw depths and the depths measured ECT by each bobbin ECT measurement technique. The results indicate that the condition employing a frequency of 280 kHz, circumferential EDM notch calibration specimen, and Vmx-based phase angle measurement exhibited the highest correlation, with a slope of 0.75, an intercept of 11.9, and a coefficient of determination (R²) of 0.46, and was therefore identified as the optimal depth sizing technique. This study overcomes the limitations of conventional ASME-based depth calibration methods by introducing an appropriate combination of calibration specimens and phase angle processing techniques that reflect the characteristics of circumferential cracks. The proposed technique can be effectively utilized as a quantitative flaw depth estimation method for the integrity assessment of feedwater heater tubes in future applications.

Eddy current testing (ECT) is periodically performed to ensure the integrity of steam generator (SG) tubes in nuclear power plants, and noise has a significant impact on the probability of detection (POD) of defects. In 2001, a leakage event at the Indian Point Unit 2 nuclear power plant in the United States was reported, where noise masked the signal of primary water stress corrosion cracking (PWSCC), highlighting the importance of effective noise management. Since then, the U.S. industry has developed noise measurement and monitoring techniques centered on the Electric Power Research Institute (EPRI), and these techniques have been incorporated into ECT inspection procedures. In this study, an EPRI-based methodology was applied to domestic nuclear power plants by introducing noise measurement and monitoring software which was pilot-tested during the ECT inspection of the Delta-53 model steam generator at Hanul Unit 1. In addition, an automated noise measurement and monitoring algorithm was developed and preliminarily validated using data from the 19th in-service inspection (ISI) of Hanul Unit 1. The POD considering noise levels was analyzed, focusing on the detection of anti-vibration bar (AVB) wear. Analysis of EPRI reference data and ECT data from the 19th and 25th inspections of Hanul Unit 1 showed that the defect depths corresponding to a 95% POD were 18%, 15%, and 10%, respectively. These results indicate that the minimum detectable defect size decreases as the noise level is reduced. In particular, the inspection with the lowest noise level (Hanul Unit 1, 25th inspection) demonstrated the best detection performance. The proposed technique provides effective data for SG integrity assessment based on ECT signal quality and is expected to contribute to improving the safety management of nuclear power plants.

The Korean nuclear industry applies ASME NQA-1 as an international quality technical standard. Recently, Korean nuclear industry signed a nuclear power plant construction contract with the Czech Republic and is required to guarantee the quality of suppliers. The need to meet the latest quality assurance requirements has also emerged in the nuclear power plant design software which is one of the supply lists. In this paper, the contents of amendments after the ASME NQA-1 2008/2009a version currently applied in Korea are reviewed, and the plan to apply the latest amendments is described.