基于响应面近似技术,对涡轮冷却叶片的气动和传热性能进行了设计优化。以冷却通道肋的位置为设计变量,采用拉丁超立方抽样在变量设计空间里选取样本点,根据样本点建立叶片计算模型,采用流-热耦合分析方法得到叶片气动与传热性能参数,拟合得到叶片壁面最高温度、平均温度和总压损失关于设计变量的四阶响应面近似模型。采用响应面模型动态修正技术,进行了回流式冷却通道的设计优化,得到了优化解,减少了总压损失,使叶片最高温度下降了24.5K,叶片壁面平均温度下降了34.4K。
In this study, an aerodynamic and heat transfer optimization design is carried out for cooling turbine blade. The locations of the ribs are determined as design variables and the Latin hypercubes technology is used to generate the experimental design points. With a coupled aerodynamic and heat transfer analysis, the blade aerodynamic and heat transfer performances are calculated at each design point, and then the quartic polynomials are employed to construct a response surface model (RSM). Based on the RSM, the optimization of the blade cooling passage is implemented to minimize the maximum temperature of the blade, the facet average temperature of the blade surface and the total pressure loss. Meanwhile, the response surface model is modified at each step of optimization. Finally, the optimization results are obtained. The maximum temperature of the blade is reduced by 24.5 K while the average blade surface temperature is reduced by 34.4 K, and the total pressure loss is also reduced.