Data-physics-model based fatigue reliability assessment methodology for high-temperature components and its application in steam turbine rotor

• Published in: Reliability engineering & system safety Vol. 241; p. 109633
• Main Authors: Gu, Hang-Hang, Wang, Run-Zi, Tang, Min-Jin, Zhang, Xian-Cheng, Tu, Shan-Tung
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.01.2024
• Subjects: Bayesian model averaging; Data-physics-model; Fatigue reliability assessment; Steam turbine rotor; Steam turbine rotor; Fatigue reliability assessment; Bayesian model averaging; Data-physics-model
• ISSN: 0951-8320; 1879-0836
• DOI: 10.1016/j.ress.2023.109633

•Constructing data-physics-model based fatigue reliability assessment framework.•Providing a physical-based HI for degradation modeling and reliability evaluation.•Achieving application in steam turbine rotor with the proposed methodology. This paper puts forward a data-physics-model based fatigue reliability assessment methodology, integrating monitoring data, physics of failure and stochastic process models, for high-temperature components. In detail, the monitored parameters are mapped into load spectrum assisted by the constructed surrogate, a physics-based health index is constructed through the damage accumulations considering uncertainties. Bayesian model averaging is utilized to combine three stochastic process models to evaluate the fatigue reliability. Furthermore, this methodology oriented to engineering application is thereafter implemented into steam turbine rotor, and the remaining useful life evaluation is presented to demonstrate the superiority of proposed method over the existing engineering method. It shows that an over conservative estimation can be avoided in application because the condition information is integrated into the fatigue modeling framework, which provides a reference to condition-based maintenance of high-temperature components.