摘要：This study proposed a design and optimization strategy for a tandem arranged cascade using the Non-dominated Sorting Genetic Algorithm(NSGA) Ⅱ multi-objective optimization algorithm and Back Propagation(BP) neural network technology. The NASA Stage 35 was employed as the initial bench mark in the present study and five geometric control parameters were working as the optimization parameters aiming to enhance the aerodynamic performance in terms of total pressure rise and efficiency. Results showed that the feasibility and capability of the proposed optimization strategy was successfully examined. In view of the fact that the initial tandem cascade(directly scaling down from NASA Stage 35) cannot guarantee the aerodynamic performance, first optimization trial was conducted to optimize the initial design. Results showed that the optimum can improve the flow quality whereas the separation on the blade is decayed or even eliminated particularly at the tip and root regions. However, compared with the initial tandem design, the enhancement in total pressure ratio(0.47%) and efficiency(1%) are too small to be noticed. Second investigation was particularly emphasizing on a high turning tandem compressor with an increment by 28°. The pressure rise and efficiency were augmented by 1.44% and 2.34%(compared to the initial tandem design), respectively. An important conclusion can be drawn that the optimization strategy is worthy to be used in high turning compressors with a considerable performance improvement.

摘要：针对导向叶片前缘大攻角引发的气动损失问题,采用数值模拟方法,开展多几何约束条件下前加载式某高负荷低压涡轮导向叶片叶型优化研究,探索导向叶片叶型关键设计参数对流动损失的影响机理,获得涡轮气动性能优化效果。结果表明:优化前叶片前缘较大正攻角,在前缘位置气流过加速,并在槽道内出现激波,导致叶片排损失增大,而优化叶型前缘进气攻角,叶片加载比较均匀,叶片出口马赫数提高1.90%,叶片气动总压损失降低19.67%,能量损失降低20.73%。