Abstract: The traditional turbine disk life prediction method does not consider the initial defects of the material, so it is impossible to predict the life of disk with defects accurately. Based on the distribution characteristics of defects produced in the different manufacturing processes, the failure probability of a turbine disk caused by defects was analyzed, focusing on the bore surface defects caused by the manufacturing process as well as the inherent material surface, sub-surface and internal defects. With the consideration of randomness of the initial defects, stress, inspection time and detectable levels, the probability of failure analysis process in the aero-engine turbine disk with defects was built. Furthermore, the analysis methods of the turbine disk with inherent defects was ameliorated, and the turbine disk was divided into four zones in which the surface zone was ulteriorly classified as surface, sub-surface and inside three parts. Finally by writing a C++ program the probability of failure analysis method was verified, and the conclusions obtained have engineering reference value. The proposed method has higher computational efficiency with the satisfied precision, and the influence about scatter of stress level, scatter of detection time and simulation times on failure probability was discussed.

Key words: Turbine disk;Damage tolerance;Probability of failure;Defect;Time of detection;Detected probability

摘要： 传统的涡轮盘寿命预测方法未考虑材料初始缺陷，因而无法对带缺陷轮盘进行较准确的寿命预测。本文以加工制造过程中产生的不同缺陷分布特征为基础，重点针对加工导致的孔表面缺陷以及与材料固有的表面、亚表面及内部缺陷，开展航空发动机涡轮盘的失效概率分析。考虑了初始缺陷、应力、检测时间、检测水平等多种随机性对涡轮盘失效概率的影响，建立了含缺陷涡轮盘的失效概率分析流程。为提高计算效率，对轮盘固有缺陷的分析方法进行改进，对轮盘重新分区，并将表层细分为表面、亚表面、内部三部分分别进行计算。通过编写C++程序分析并验证了含缺陷轮盘失效概率分析方法的可行性，获得的结论具有工程应用参考价值。本文方法在满足一定精度的同时具有较高的计算效率，并对应力水平、检测时间的分散性和模拟次数等对失效概率的影响进行了讨论。

关键词: 涡轮盘;损伤容限;失效概率;缺陷;检测时间;检出概率