A data-driven roadmap for creep-fatigue reliability assessment and its implementation in low-pressure turbine disk at elevated temperatures

• DOI: 10.1016/j.ress.2022.108523
• Publication Date: 2022-04-17T21:17:06Z
• Authors: Run-Zi Wang; Hang-Hang Gu; Shun-Peng Zhu; Kai-Shang Li; Ji Wang; Xiao-Wei Wang; Miura Hideo; Xian-Cheng Zhang; Shan-Tung Tu
• Keywords: Service life

High-reliability life design process not only can ensure system safety in service, but also provide scientific management during the maintenance period. The objective of present work is to develop a roadmap for creep-fatigue reliability assessment. Material-level data accumulations and theoretical foundations including constitutive multi-axial damage models are introduced. Afterwards, low-pressure turbine disk under typical load waveform applied as case study demonstrate how perform assessment by using this practice. Precise weakness hotspots identified at mortise joint disk. Based on hotspot-based strategy, it found that surrogate model assisted an optimized machine learning method enhances significantly computational efficiency. Accordingly, probabilistic with considering multi-sources uncertainty obeys lognormal distributions. In aspect failure probability analysis, current interaction diagram gives conservative assessments excellent universality compared traditional ones mainly used low cycle fatigue field. Last least, evaluation discussed comprehensively consider potential occurrence averaged hot region instead single hotspot.