提出了一种利用复合疲劳试验和外场故障数据反推涡轮叶片实际振动应力的方法.该方法针对与故障叶片同批次的叶片,开展数个振动应力水平下的单件试验和某一特定振动应力水平下的成组试验,利用极大似然法推导出叶片的概率-应力一寿命曲线(P—S—N);最后基于99.87%存活率下的概率-应力-寿命曲线(P99.87%-S—N),结合叶片的外场故障统计结果,反推出叶片实际工作中振动应力的范围和可能的最大振动应力.
An inverse method based on combined high-and-low cycle fatigue (HCF/ LCF) tests and the outfield practical failure data was developed to estimate the vibration stress of the turbine blade in working condition. All the test specimens of the turbine blades owned the same production procedures as the outfield failure blades. Two groups of HCF/ LCF tests were performed, one was median life test at six vibration stress levels, and the other was life dispersion test at a given vibration stress level. Then, P-S-N curves were extrapolated with the maximum likelihood method. Finally, vibration stress of the turbine blade in working condition was determined by the P99.87%-S-N curve combined with the statistical life of the outfield practical failure data.