Efficient generation method and mechanical property prediction of pore-embedded representative volume element for fiber-reinforced composites

In the manufacturing of carbon fiber-reinforced composites, defects such as porosity tend to occur within the matrix, mainly caused by inappropriate process parameters, uneven raw material distribution, solvent evaporation, or air entrapment. Extensive research has been devoted to porosity effects, and finite element analysis has become the dominant method for such investigations. Accordingly, the accurate and efficient modeling of representative volume elements (RVEs) containing pores is essential for relevant studies. This paper proposes an efficient modeling method for pore-embedded representative volume elements (PeRVEs), which allows fast generation of models with diverse pore characteristics. The method features a clear principle and convenient implementation. Based on this approach, a customized plugin for ABAQUS is developed. Key factors affecting modeling efficiency are analyzed, and the equivalent mechanical properties of unidirectional composites are predicted to validate the model reliability. The effects of porosity on the strength of unidirectional composites are also examined. Results show that porosity significantly influences the transverse tensile and compressive strengths of the fibers. The proposed method offers reliable support for microscopic modeling in computational micromechanics.