Nonlinear dynamic probabilistic design of turbine disk-radial deformation using extremum response surface method-based support vector machine of regression

2014 ◽  
Vol 229 (2) ◽  
pp. 290-300 ◽  
Author(s):  
Cheng-Wei Fei ◽  
Wen-Zhong Tang ◽  
Guang-Chen Bai
Keyword(s):  
Support Vector Machine ◽  
Response Surface ◽  
Response Surface Method ◽  
Nonlinear Dynamic ◽  
Turbine Disk ◽  
Support Vector ◽  
Radial Deformation ◽  
Probabilistic Design ◽  
Surface Method

Nonlinear Dynamic Probabilistic Analysis for Turbine Casing Radial Deformation Using Extremum Response Surface Method Based on Support Vector Machine

2013 ◽  
Vol 8 (4) ◽  
Author(s):  
Chengwei Fei ◽  
Guangchen Bai
Keyword(s):  
Support Vector Machine ◽  
Response Surface ◽  
Response Surface Method ◽  
Probabilistic Analysis ◽  
Nonlinear Dynamic ◽  
Nonlinear Material ◽  
Support Vector ◽  
Radial Deformation ◽  
Surface Method ◽  
Turbine Casing

To improve the computational efficiency of nonlinear dynamic probabilistic analysis for aeroengine typical components, an extremum response surface method based on the support vector machine (SVM ERSM) was proposed in this paper. The basic principle was introduced and the mathematical model was established for the SVM ERSM. The probabilistic analysis of turbine casing radial deformation was taken as an example to validate the SVM ERSM considering the influences of nonlinear material property and dynamic heat loads. The results of probabilistic analysis imply that the distribution features of random parameters and the major factors are gained for more accurate the design of casing radial deformation. The SVM ERSM offers a feasible and valid method, which possesses high efficiency and high precision in the nonlinear dynamic probabilistic analysis. Moreover, the SVM ERSM is promising to provide an useful insight for casing dynamic optimal design and the blade-tip clearance control of aeroengine high pressure turbine.

A dynamic probabilistic design method for blade-tip radial running clearance of aeroengine high-pressure turbine

2014 ◽  
Vol 229 (10) ◽  
pp. 1861-1872 ◽  
Author(s):  
Cheng-Wei Fei ◽  
Wen-Zhong Tang ◽  
Guang-chen Bai ◽  
Zhi-Ying Chen
Keyword(s):  
High Pressure ◽  
Response Surface ◽  
Response Surface Method ◽  
Probabilistic Analysis ◽  
High Reliability ◽  
Probabilistic Design ◽  
High Pressure Turbine ◽  
Analysis And Design ◽  
Surface Method ◽  
Blade Tip

Around the engineering background of the probabilistic design of high-pressure turbine (HPT) blade-tip radial running clearance (BTRRC) which conduces to the high-performance and high-reliability of aeroengine, a distributed collaborative extremum response surface method (DCERSM) was proposed for the dynamic probabilistic analysis of turbomachinery. On the basis of investigating extremum response surface method (ERSM), the mathematical model of DCERSM was established. The DCERSM was applied to the dynamic probabilistic analysis of BTRRC. The results show that the blade-tip radial static clearance δ = 1.82 mm is advisable synthetically considering the reliability and efficiency of gas turbine. As revealed by the comparison of three methods (DCERSM, ERSM, and Monte Carlo method), the DCERSM reshapes the possibility of the probabilistic analysis for turbomachinery and improves the computational efficiency while preserving computational accuracy. The DCERSM offers a useful insight for BTRRC dynamic probabilistic analysis and optimization. The present study enrichs mechanical reliability analysis and design theory.

Sign in / Sign up

Export Citation Format

Share Document