论文部分内容阅读
为了正确评估加工误差的影响,提出了一种考虑加工误差大小、分布随机性的数值研究方法以实现叶片性能不确定性分析。采用NURBS(非均匀有理B样条)方法实现叶型参数化,通过检测及统计手段获取误差真实分布,运用蒙特卡洛模拟生成随机样本,并训练Kriging代理模型进行不确定性分析,得到了叶片加工误差对气动性能的影响。结果表明:加工误差对气动性能的影响与所处工况有关,最高使得损失增加0.853%;叶型性能对于吸力面表面型线更为敏感,在加工及检测过程应给予更高要求。
In order to correctly evaluate the effect of machining error, a numerical method considering the size of machining errors and randomness of the distribution is proposed to realize the uncertainty analysis of the blade performance. NURBS (non-uniform rational B-spline) method was used to realize the parameterization of leaf shape. The error distribution was obtained by means of detection and statistics. The Monte Carlo simulation was used to generate random samples. The Kriging model was trained for uncertainty analysis. Leaf shape Effect of machining error on aerodynamic performance. The results show that the influence of machining error on the aerodynamic performance is related to the working conditions, and the maximum loss is increased by 0.853%. The aerodynamic performance of the blade is more sensitive to the surface profile of the suction surface and the machining and testing process should be given higher requirements.