Abstract: The quality of resin matrix fiber reinforced composites is closely related to the resin flow impregnation during the manufacturing process. As the cornerstone of meso scale and macro scale, the micro scale fiber structure and micro permeability have a direct impact on the flow impregnation of resin. However, it is difficult to effectively carry out experiments at the micro scale. Thus, numerical simulation has become a more effective research method. In this paper, Monte Carlo stochastic algorithm is applied to establish the micro scale models of the random arrangement of filaments in the fiber tow and the mesoscopic scale structure intra and inter the fiber tows. The permeability of the inner structure of the fiber tow is predicted based on the finite element method, and the phase field method is creatively used to simulate the resin flow considering the surface tension and capillary force. The results show that the prediction accuracy of the random structure permeability prediction method of fiber filament established in this paper is significantly higher than that of the traditional empirical formula. At the same time, the pore formation process in the micro fiber bundle is simulated, which is of great significance to expound the resin flow impregnation process and the generation mechanism of micro bubbles from the micro scale system.

Key words: fiber micro scale, stochastic algorithm, permeability, resin flow, numerical simulation, composites