Fatigue strength reliability assessment of turbo-fan blades by Kriging-based distributed collaborative response surface method - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Ekspolatacja i Niezawodnosc - Maintenance and Reliability Année : 2019

Fatigue strength reliability assessment of turbo-fan blades by Kriging-based distributed collaborative response surface method

(1, 2) , (1) , (3, 2, 4, 5) , (1)
1
2
3
4
5

Résumé

Fatigue crack propagation affects the operational reliability of engine turbo-fan blades. In this article, we integrate a Kriging regression model and a distributed collaborative response surface method (DCRSM) for the reliability assessment of turbo-fan blades, considering the relevant uncertainty. Following a series of deterministic analyses, such as steady-state aerodynamic analysis, harmonic response analysis and Campbell diagram, and based on the assumption that vibration stress is mainly from aerodynamic load, the fatigue strength is calculated for turbo-fan blades under coupling aerodynamic forces, according to a modified Goodman curve of titanium-alloy. Giving consideration to the uncertainty of the resonance frequencies and material properties, the fatigue strength of the turbo-fan blade is evaluated, including probabilistic analysis and sensitivity analysis. In the case study analyzed, the conclusions are that the fatigue strength reliability reaches 96.808% with confidence level of 0.95 for the turbo-fan blade under the coupling aerodynamic forces, and the first three-order resonant frequencies are found to have important influence on the fatigue performance of turbo-fan blades.

Dates et versions

hal-02433490 , version 1 (09-01-2020)

Identifiants

Citer

Hai-Feng Gao, Anjenq Wang, Enrico Zio, Wei Ma. Fatigue strength reliability assessment of turbo-fan blades by Kriging-based distributed collaborative response surface method. Ekspolatacja i Niezawodnosc - Maintenance and Reliability, 2019, 21 (3), pp.530-538. ⟨10.17531/ein.2019.3.20⟩. ⟨hal-02433490⟩
49 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook Twitter LinkedIn More