高温叶片换热性能不确定性量化  

Uncertainty Quantification of Heat Transfer Performance of High Temperature Blade

作　　者：宋英杰[1] 聂俊[1] 郭振东[1] 宋立明[1] 李军[1] 丰镇平[1] 

机构地区：[1]西安交通大学叶轮机械研究所,西安710049

出　　处：《工程热物理学报》2015年第8期1666-1671,共6页Journal of Engineering Thermophysics

基　　金：国家自然科学基金资助项目(No.51106123)

摘　　要：结合共轭换热分析方法和基于拉丁超立方采样与Kriging近似模型的蒙特卡罗方法,提出了高温叶片换热性能不确定性量化方法。在量化主流进口总压、总温、湍流强度以及冷气进口总压四个运行参数的不确定性的基础上,对高温叶片MarkII进行了换热性能不确定性量化。分析了不确定性工况参数对叶片最高温度和最大温度梯度两个关键随机输出参数的影响,并通过设定最高温度和最大温度梯度的临界值,进行了可靠性分析。计算结果表明,不确定性参数可导致叶片最高温度增加42.3 K,最大温度梯度增加2.20 K·mm^(-1);最高温度下和最大温度梯度下的可靠度分别为94.2%和97.0%;主流进口总温不确定性对叶片换热性能和寿命影响最大。Monte Carlo simulation method based on Kriging response surface and Latin Hypercube sampling was developed,coupled with conjugate heat transfer（CHT） analysis,an uncertainty quantification method for heat transfer performance of high temperature blade was proposed.Taking four operating condition uncertainty parameters into account,uncertainty quantification of heat transfer performance of Mark II blade was carried out.Statistics of two stochastic output quantities of interest was calculated,including maximum metal temperature and maximum temperature gradient,and reliability analysis was performed based on respective critical values.The results indicate that the maximum increase of the maximum temperature and maximum temperature gradient could reach as much as 42.3 K and 2.20 K-mm（-10）,respectively.The reliability in term of maximum temperature and maximum temperature gradient is 94.2%and 97.0%,respectively.The results also show that the uncertain parameter which mostly affects heat transfer performance is the inlet total temperature.

关 键 词：不确定性量化 共轭换热分析 蒙特卡罗模拟 拉丁超立方采样 Kriging响应面 

分 类 号：O354[理学—流体力学]