基于分解策略的涡轮叶片可靠性及多学科设计优化

将反一阶可靠性分析方法与多学科可行方法相结合,提出了一种适用于涡轮叶片复杂结构的可靠性及多学科设计优化方法.在优化过程中使用Kriging近似模型并不断提高模型精度,解决了多学科可行方法反复调用仿真程序进行多学科分析,计算量较大的问题.该方法将可靠性分析与多学科优化过程分离,提高了优化计算效率.以某型涡轮叶片的设计优化为例,对该方法进行了验证并与传统双循环方法进行了对比.结果表明,优化结果满足可靠性的要求,与双循环方法相比优化效率提高63.8%,证明了该方法在工程应用中的可行性和有效性.

Reliability and multidisciplinary design optimization for turbine blade based on decoupled method

A suitable reliability-based multidisciplinary design optimization(RBMDO) framework for turbine blade was proposed by combining inverse first order reliability analysis method(inverse FORM) with multidisciplinary feasible(MDF) method.The Kriging approximate model with updating function was introduced to reduce the computational cost of MDF caused by complex structure.The proposed method is efficient since it can decouple reliability analysis process from optimization process.A turbine blade design optimization was employed to illustrate the feasibility and efficiency of the proposed method.It shows that the optimal results can satisfy the required reliability and the efficiency is improved 63.8% by the proposed method as compared with commonly used double-loop based RBMDO method.It is feasible and efficient to apply this method to engineering design.