反应堆运行时，锆合金包壳会承载一定循环载荷，在设计过程中必须要考虑包壳材料的疲劳性能，防止疲劳失效。为了更准确地拟合和预测锆合金的疲劳性能，以两种不同热处理状态锆合金疲劳试验数据为数据基础，采用概率权重矩方法得到三参数威布尔分布模型拟合得到可靠度-应力-寿命（R-S-N）曲线，并与基于传统Basquin模型的R-S-N曲线拟合结果进行对比验证，发现基于三参数威布尔分布的S-N曲线具备更高的拟合准确度，显著优于传统方法拟合结果，对锆合金疲劳性能分析与预测具有良好的适用性。

During the operation of nuclear reactors, zirconium alloy claddings are subjected to significant cyclic stresses, necessitating a comprehensive assessment of the material’s fatigue properties to prevent structural failure. This study focuses on enhancing the accuracy of predictive models for the fatigue performance of zirconium alloys by employing fatigue test data from alloys in two distinct heat-treated states. A three-parameter Weibull distribution model, formulated using the probabilistic weighted moments method, was developed to construct Reliability-Stress-Number of cycles (R-S-N) curves. These curves were rigorously compared and validated against the R-S-N curves obtained using the traditional Basquin model. The findings indicate that the S-N curves, derived from the three-parameter Weibull distribution, demonstrate superior fitting accuracy and markedly surpass the performance of those obtained through conventional methodologies. This advanced modeling approach thus proves to be highly effective and applicable for the detailed analysis and prediction of fatigue properties in zirconium alloys, offering significant implications for the design and safety of nuclear reactors.

TG 146.414；TL 341

四川省自然科学基金(Grant No.2023NSFSC0979)；国家重点研发计划（Grant No.2022YFB1902402）