粗糙叶片真实形貌条件下的压气机内部不确定性流动分析方法研究

Compressors in gas turbines with service inevitably suffer from blade fouling, corrosion and erosion, which has been acknowledged as the leading cause of gas turbine performance degradation. It is thus of great urgency for intensive research into the compressor performance deterioration and the associated flow mechanisms due to blade fouling, corrosion and erosion. In this project, a new and generic geometry parameterization model will be built by means of experimental measurement and theoretical analysis, in order to characterize the non-uniform, randomly accidented and multi-scaled surface morphology of the real rough blade. A high-dimensional stochastic CFD method with high efficiency will be developed to overcome the so-called ‘curse of dimensionality’, which is the bottleneck problem of the high-dimensional uncertainty analysis. Finally, an uncertainty analysis method for the internal flow of compressors with service will be constructed, whereby the statistic natures of compressor cascade performance and flow field under geometry uncertainties of the real rough blade are to be revealed. This research will provide scientific and technical foundations for understanding the uncertain flow phenomena in gas turbine compressors with service, for improving the service performance of gas turbines and their major components, and for promoting the associated robust optimization design and control methods.

实际服役环境下的燃气轮机压气机部件不可避免地会遭遇叶片的积垢、腐蚀及磨损，而由此造成的压气机性能恶化已被公认为引起燃机系统效能衰退甚至失效的首要诱因。为此，深入研究叶片积垢、腐蚀及磨损条件下的压气机气动性能恶化现象及流动机理迫在眉睫。本项目拟通过实验测量与理论建模，建立起压气机粗糙叶片真实形貌的全新的通用几何参数化表征模型，以囊括因积垢、腐蚀及磨损造成的叶片表面粗糙度空间分布不均匀、随机凸凹及尺度跨度大的真实形貌特征；拟发展高维随机CFD方法及其高效化计算途径，解决或缓解高维不确定性分析中遇到的“维数灾难”瓶颈问题；而后构建出一套粗糙叶片真实形貌条件下的压气机内部不确定性流动分析方法，首次揭示粗糙叶片真实形貌几何不确定性对压气机叶栅特性及内部流场的统计影响规律。项目为深刻认识实际服役环境下压气机内部的不确定性流动现象、提升燃气轮机及其关键部件的服役性能和鲁棒优化设计与控制水平提供科技基础。

项目面向高维高效不确定性分析的学术前沿，围绕任务书设定的研究目标，针对实际服役环境下的压气机叶片开展不确定性建模和不确定性流动分析研究。在不确定性建模方面，基于傅里叶级数理论，建立了表征二维粗糙叶片表面任意凸起或凹陷的几何参数化模型；并将Hicks-Henne函数从二维平面延拓到三维空间、从低阶拓展至高阶，建立了表征三维叶片表面粗糙形貌的几何参数化模型。在不确定性流动分析方面，采用计算流体力学方法，先后研究了平面叶栅和跨音速压气机转子Rotor37在粗糙叶片条件下的气动性能和内部流场；为克服高维不确定性分析中的“维数灾”，提出了一种分层降维近似模型方法，并将其成功应用至粗糙形貌条件下的平面叶栅气动性能和内部流动不确定性分析中，结果表明，相对于叶片吸力面表面局部凹凸形貌的空间疏密程度，叶栅总压损失系数对叶片吸力面表面局部凹凸形貌的尺寸大小更为敏感，且叶栅马赫数分布的一阶和二阶模态分别主导其均值和标准差的分布。研究成果为深刻认识实际服役环境下压气机内部的不确定性流动现象、提升燃气轮机及其关键部件的服役性能和鲁棒优化设计与控制水平提供科技基础。