Machine learning compliance calibration for local fatigue energy release rate calculations in multi-directional laminates with fiber bridging

Abstract

This study presents a numerical analysis framework of multidirectional (MD) composite laminate under mode I fatigue loading, with calibration to account for fiber bridging. Both local and global methods for calculating the energy release rate are compared. Seven unidirectional (UD) and MD layups were tested, revealing differences in fatigue resistance based on fiber orientation and initial delamination length. Fiber bridging, where fibers stretch across separating plies, was found to enhance toughness. The results show that fiber bridging intensity varies with fiber orientation, with minimal stiffening for the 0°//0° interface and significant stiffening for 90°//90°, 0°//45°, and 0°//90° interfaces.