A New Adaptive Response Surface Model for Reliability Analysis of 2.5D C/SiC Composite Turbine Blade

作者：Chang, Yaning; Sun, Zhigang^*; Sun, Weiyi; Song, Yingdong

来源：Applied Composite Materials, 2018, 25(5): 1075-1091.

DOI：10.1007/s10443-017-9652-2

• Reliability analysis
• 2.5D C/SiC composites
• Adaptive response surface approach
• Gaussian regression
• Latin hypercube sampling

In order to calculate the failure probability of complex structures such as a 2.5D/SiC composites turbine blade and improve the structure safety, a new adaptive model of Response Surface (RS) analysis has been developed in this paper, which can improve the computational efficiency of structural failure problem while ensure the accuracy. The Gaussian Process Regression (GPR) theory was used to establish the RS and reconstruct the performance function of structure. And, an Adaptive Latin hypercube Sampling (ALHS) strategy was adopted in the process of establishing and correcting the RS. Finally the Direct Simulation Monte Carlo(DSMC)was utilized to calculate the failure probability of the performance function replacing the complex structure. Two numerical examples were calculated to validate the accuracy and computational efficiency of the proposed method. Additionally the finite element stress analysis results of 2.5D C/SiC composite turbine blade were used to structural reliability analysis by the proposed method. The approach in this paper provides a new way to evaluate the risk of the complex structures.

• 出版日期2018-10
• 单位南京航空航天大学