A method for predicting the mechanical properties of short fiber reinforced polymer composites based on fiber orientation distribution image processing technique

Short fiber reinforced polymer composites (SFRPC) possess complex microscale structures, and understanding the fiber orientation distribution (FOD) within the SFRPC is a prerequisite for mechanical modeling of short fiber composites. However, the statistical analysis of fiber orientation requires collecting a large amount of fiber orientation information, and traditional manual annotation and retrieval of micrographs are costly and time-consuming, making it challenging to ensure both statistical efficiency and accuracy. In this study, image analysis algorithms were employed to capture geometric features of fiber cross-sections, and a corresponding image processing technique for FOD was developed, enabling the rapid statistical analysis of FOD information. The reasonable range of key parameters in image analysis algorithms was explored, and microstructure characterization was conducted on short glass fiber-reinforced and short carbon fiber-reinforced polyetherimide (SGF/PEI and SCF/PEI) composites fabricated using extrusion and injection molding processes. The statistical fiber orientation information was then incorporated into the laminate analogy approach (LAA) and Fu-Lauke model frameworks to predict the modulus and strength of the two composites with different volume fractions. The predicted results exhibited good agreement with finite element simulation results and experimental tensile test data. By combining the image processing technique for FOD with the prediction methods for composite mechanical properties, this study is helpful to more efficient and accurate understanding of the structure-property relationship of short fiber reinforced composites, providing valuable guidance for composite structural design.