Research on Residual Life Prediction Method of Composites Based on Equivalent Number of Cycles Conversion

Qiang Ma,Zihao Feng,Huidong Ma,Zongwen An,Shilong Zeng,Xuezong Bai

DOI: https://doi.org/10.1007/s11668-024-01875-0

2024-01-01

Journal of Failure Analysis and Prevention

Abstract:It is crucial to propose an accurate fatigue life prediction method to ensure the structural integrity and operational reliability of engineering components in real-world service conditions. This paper considers how the uncertainty of fatigue failure in composite materials under variable amplitude loading conditions affects material fatigue life. It proposes an improved method to predict the residual fatigue life of a composite material by converting the number of loading cycles under different loading sequences. Firstly, the method suggests that the number of loading cycles for the same composite material, under variable amplitude loading conditions and at different stress levels, conforms to the assumptions of the Weibull cumulative distribution and the principle of probabilistic consistency. Using this assumption, a transformation model is established for the number of cyclic loading cycles in composite materials under variable amplitude loading conditions. Specifically, the model can convert the number of loading cycles under various loading sequences into an equivalent number of cycles under a uniform loading condition. Next, the residual fatigue life of the specimen at the final stress level was determined by analyzing fatigue test data under variable amplitude loading conditions. Finally, the proposed method is validated and compared using fatigue test data from the literature, which includes composites tested under two and three variable amplitude loading conditions. Validation results indicate that the proposed method for predicting residual fatigue life offers more accurate predictions, with all predicted life factors falling within ± 1.2 life factors. Additionally, compared to other existing models for predicting residual life, the proposed method effectively captures the impact of randomness and uncertainty in the fatigue failure process of composite materials using probabilistic statistical theory. This offers a more valuable reference for predicting fatigue life under variable amplitude load.