Li Zhijie, Peng Fei, Zhao Jian, Guo Sujuan, Hu Lefei, Gong Yu
Key Laboratory of AI-Aided Airworthiness of Civil Aircraft Structures, School of Aerospace Engineering and Applied Mechanics, Tongji University, Shanghai 200092, China.
Beijing Institute of Astronautical Systems Engineering, Beijing 100076, China.
Materials (Basel). 2025 Aug 4;18(15):3667. doi: 10.3390/ma18153667.
Fiber-reinforced resin matrix composites are generally composed of fibers and matrix with significantly different properties, which are non-uniform and anisotropic in nature. Macro-failure criteria generally view composite plies as a uniform whole and do not accurately reflect fiber- and matrix-scale failures. In this study, the interface phase effect between fiber and matrix has been introduced into the frame of trans-scale analysis to better model the failure process, and the equivalent mechanical property characterization model of the interface phase has also been established. Combined with the macro-micro-strain transfer method, the trans-scale correlation of the mechanical response of the composite laminates between the macro scale and the fiber, matrix and interface micro scale has been achieved. Based on the micro-scale failure criterion and the stiffness reduction strategy, the trans-scale failure analysis method of composite materials incorporating the interface phase effect has been developed, which can simultaneously predict the failure modes of the matrix, fiber and interface phase. A numerical implementation of the developed trans-scale failure analysis method considering interface phase was carried out using the Python and Abaqus 2020 joint simulation technique. Case studies were carried out for three material systems, and the prediction data of the developed trans-scale failure analysis methodology incorporating interface phase effects for composite materials, the prediction data of the Linde failure criterion and the experimental data were compared. The comparison with experimental data confirms that this method has good prediction accuracy, and compared with the Linde and Hashin failure methods, only it can predict the failure mode of the fiber-matrix interface. The case analysis shows that its prediction accuracy has been improved by about 2-3%.