中文摘要：

To reasonably design the blade-tip radial running clearance(BTRRC) of high pressure turbine and improve the performance and reliability of gas turbine, the multi-object multi-discipline reliability sensitivity analysis of BTRRC was accomplished from a probabilistic prospective by considering nonlinear material attributes and dynamic loads. Firstly, multiply response surface model(MRSM) was proposed and the mathematical model of this method was established based on quadratic function. Secondly, the BTRRC was decomposed into three sub-components(turbine disk, blade and casing), and then the single response surface functions(SRSFs) of three structures were built in line with the basic idea of MRSM. Thirdly, the response surface function(MRSM) of BTRRC was reshaped by coordinating SRSFs. From the analysis, it is acquired to probabilistic distribution characteristics of input-output variables, failure probabilities of blade-tip clearance under different static blade-tip clearances δ and major factors impacting BTRRC. Considering the reliability and efficiency of gas turbine, δ=1.87 mm is an optimally acceptable option for rational BTRRC. Through the comparison of three analysis methods(Monte Carlo method, traditional response surface method and MRSM), the results show that MRSM has higher accuracy and higher efficiency in reliability sensitivity analysis of BTRRC. These strengths are likely to become more prominent with the increasing times of simulations. The present study offers an effective and promising approach for reliability sensitivity analysis and optimal design of complex dynamic assembly relationship.

英文摘要：

To reasonably design the blade-tip radial running clearance（BTRRC） of high pressure turbine and improve the performance and reliability of gas turbine, the multi-object multi-discipline reliability sensitivity analysis of BTRRC was accomplished from a probabilistic prospective by considering nonlinear material attributes and dynamic loads. Firstly, multiply response surface model（MRSM） was proposed and the mathematical model of this method was established based on quadratic function. Secondly, the BTRRC was decomposed into three sub-components（turbine disk, blade and casing）, and then the single response surface functions（SRSFs） of three structures were built in line with the basic idea of MRSM. Thirdly, the response surface function（MRSM） of BTRRC was reshaped by coordinating SRSFs. From the analysis, it is acquired to probabilistic distribution characteristics of input-output variables, failure probabilities of blade-tip clearance under different static blade-tip clearances δ and major factors impacting BTRRC. Considering the reliability and efficiency of gas turbine, δ=1.87 mm is an optimally acceptable option for rational BTRRC. Through the comparison of three analysis methods（Monte Carlo method, traditional response surface method and MRSM）, the results show that MRSM has higher accuracy and higher efficiency in reliability sensitivity analysis of BTRRC. These strengths are likely to become more prominent with the increasing times of simulations. The present study offers an effective and promising approach for reliability sensitivity analysis and optimal design of complex dynamic assembly relationship.