TITLE:
Fatigue Life Prediction of Vehicle Rubber Elastic Support Components Based on Physics-Informed Neural Networks
AUTHORS:
Shen Liu, Fei Meng
KEYWORDS:
Data-Driven Model, Fatigue Life, PINN Model, Rubber Isolators, Stiffness
JOURNAL NAME:
Open Journal of Applied Sciences,
Vol.16 No.1,
January
19,
2026
ABSTRACT: Rubber is widely used in automotive vibration isolation systems due to its excellent mechanical properties and durability. However, elastomeric support components tend to experience performance degradation under high-temperature and cyclic loading conditions. Consequently, predicting the fatigue life of these components has been a critical issue in structural reliability research. Due to the strong nonlinearity and significant energy dissipation characteristics of rubber, traditional empirical models are often limited by small sample sizes and lack physical interpretability, making it difficult to accurately describe the damage evolution process under complex operating conditions. To enhance the reliability of fatigue predictions, this study proposes a modeling approach based on Physics-Informed Neural Networks (PINN). By integrating physical modeling with data-driven techniques and embedding damage evolution equations into the learning framework, the model not only adheres to physical laws but also fits the measured stiffness degradation curves, even with limited experimental data. The results show that this method effectively reconstructs the material’s fatigue damage process and reliably predicts its fatigue life. The proposed PINN framework offers an efficient and physically consistent approach for evaluating the fatigue life of rubber-based support components.