Structural reliability analysis and uncertainties‐based collaborative design and optimization of turbine blades using surrogate model

• Published in: Fatigue & fracture of engineering materials & structures 2019-06, Vol.42 (6), p.1219-1227
• Main Authors: Meng, Debiao, Yang, Shunqi, Zhang, Yu, Zhu, Shun‐Peng
• Format: Article
• Language: English
• Subjects: Collaboration; Computer simulation; Design analysis; Design optimization; Mathematical models; multidisciplinary design and optimization; Reliability analysis; Reliability engineering; Response surface methodology; saddlepoint approximation reliability analysis; Structural reliability; surrogate models; turbine blade component; Turbine blades; turbine equipment; Turbines; Uncertainty analysis
• ISSN: 8756-758X; 1460-2695
• DOI: 10.1111/ffe.12906

In power and energy systems, both the aerodynamic performance and the structure reliability of turbine equipment are affected by utilized blades. In general, the design process of blade is high dimensional and nonlinear. Different coupled disciplines are also involved during this process. Moreover, unavoidable uncertainties are transported and accumulated between these coupled disciplines, which may cause turbine equipment to be unsafe. In this study, a saddlepoint approximation reliability analysis method is introduced and combined with collaborative optimization method to address the above challenge. During the above reliability analysis and design optimization process, surrogate models are utilized to alleviate the computational burden for uncertainties‐based multidisciplinary design and optimization problems. Smooth response surfaces of the performance of turbine blades are constructed instead of expensively time‐consuming simulations. A turbine blade design problem is solved here to validate the effectiveness and show the utilization of the given approach.