全局叶型扭转角误差对轴流压气机性能影响的不确定性量化分析方法

To investigate the impact of global twist angle errors on Performance Parameters and flow ficld structure of transonic axial compressor blades, an uncertainty quantification process and method are being proposed for studying Rotor37. A thrcc-dimensional reduced-order model of geometric uncertainty is being constructed based on the principal componcnt analysis method. By selecting three-dimensional normal distribution variables in a non-intrusive chaotic polynomial, numerical methods are used to solvc the füll three-dimensional Rcynolds-avcraged Navier-Stokcs equations at required sample points to construct a non-intrusive chaotic polynomial for evaluating Performance responses with a large sample size. The findings rcveal that the Statistical results of Performance Parameters show a normal distribution with an unaffected mean level but with around 1 % fluctuation. Through Spearman corrclation analysis, critical spanwise section positions influencing the flow field are identified, indicating the strongest negative correlation between efficiency and twist angle errors in the 70% to 100% blade height ränge. This emphasizcs the nccd to improve machining accuracy in this region. Analysis of flow field fluctuations under Statistical significancc and flow field characteristics of bladcs with extreme efficiency reveals the mechanism of twist angle errors on Performance. The results show that the most severe flow field fluctuations occur at the tip clearance leakage flow and shock wave positions, whcre twist angle errors affeet both positions and the intensity of their mutual interference, thus influencing downstream blockage and ultimately affecting efficiency. It is evident that the influence of global twist angle errors on blade Performance and flow field cannot be ignored, and should bc carcfully considered in robust design based on flow field characteristics.