考虑几何分散性的涡轮盘寿命概率分析

针对考虑几何分散性的涡轮盘低循环疲劳（LCF）寿命概率分析中几何参数多、几何随机变量难确定、分布特征获取困难、模型需自动更新及计算成本高的问题，提出几何分散性的概率处理方法:采用试验设计方法对涡轮盘结构所有几何参数进行灵敏度分析,筛选出对应力影响较大的关键几何参数作为随机变量,使用K-S(Kolmogorov Smirnov)方法确定其分布类型和特征参数,最后建立代理模型进行Monte Carlo概率分析.基于此方法,开发出了涡轮盘概率分析系统,在该系统中筛选得到某发动机GH720Li涡轮盘内径、外径、盘缘厚度3个结构参数作为几何随机变量,完成对LCF寿命的概率分析工作得到寿命-可靠度分布曲线.分析结果表明涡轮盘外径对LCF寿命有较大影响. 

Probabilistic analysis of turbine disk fatigue life considering geometric uncertainties

For the issues of a large number of geometric variables, hard to select geometric random variables and determine their probability distributions, rebuilding the finite element model in an automated fashion, and huge numerical computation comsuption when conducting the LCF (low cycle fatigue) life prediction considering geometric uncertainties of turbine disk. A geometric uncertainty probabilistic processing method was proposed. First, the design of experiments method was employed to determine key geometric parameters, which would be treated as random variables for LCF prediction. Second, the K S (Kolmogorov Smirnov) method was adopted to determine the probability distributions of random variables. Third, a surrogate model was built to conduct Monte Carlo simulation. On this basis, turbine disk probabilistic analysis system was developed. The LCF life probabilistic analysis work of a GH720Li turbine engine disk was completed under the turbine disk probability analysis system, by which three key geometric parameters (inner diameter, outer diameter, rim thickness) and life reliability curve of the turbine disk were obtained. Analysis results shows that the outer diameter of turbine disk has a strong impact on the LCF life.