Acoustic emission-driven fatigue damage evolution equation and life prediction of composite laminates

Conventional acoustic emission (AE) establishes correlations between signals and damage, often relying on machine learning or threshold-based approaches. These methods either lack quantitative rigor or depend on black-box models with limited interpretability. To address these limitations, this study leverages AE signals to develop a physics-informed damage evolution equation, capturing the underlying physical mechanisms governing fatigue damage in composite laminates. By integrating AE signal analysis with a phenomenological framework, the proposed model extracts damage state from AE data and incorporates material-specific physical parameters to accurately describe damage progression and predict fatigue life. This approach characterizes the fatigue damage evolution using AE cumulative energy, effectively reconstructing the three critical stages of fatigue: primary damage, steady-state damage, and accelerated damage. AE data from plain-woven glass fiber/cyanate composite laminate (GFRP) and multidirectional symmetric carbon fiber CCF800H/AC531 laminates (CFRP) with open-hole subjected to tension–tension loading were utilized for validation. The results showed that the predicted fatigue life for GFRP and CFRP falls within 2 and 2.5 times their respective error bands. These findings underscore the potential of this physics-guided AE-based methodology for applications in structural health monitoring and fatigue life prediction of composite materials.