基于纤维取向分布图像处理技术的短纤维增强聚合物基复合材料力学性能预测方法

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

  • 摘要: 短纤维增强聚合物基复合材料(Short fiber reinforced polymer composites,SFRPC)具有复杂的细观结构,掌握纤维取向分布(Fiber orientation distribution,FOD)规律是短纤维复合材料力学建模的前提。然而,由于纤维取向统计需要收集大量的纤维信息,通过传统手动标注读取显微图像的方式人工成本高且耗时长,统计效率与精度均难以保证。本文利用图像分析算法捕获纤维截面几何特征,发展了相应的纤维取向分布图像处理技术,实现了FOD信息的快速统计。探究了图像分析算法中关键参数的合理取值范围,并针对挤出注塑成型工艺制备的短玻纤增强和短碳纤增强聚醚酰亚胺复合材料(SGF/PEI 和SCF/PEI)进行微观结构表征,将统计的纤维状态信息传递至类层合板(Laminate analogy approach,LAA)与Fu-Lauke模型框架,进而预测了不同体积分数下两种复合材料的模量与强度,预测结果与有限元模拟结果、拉伸试验测试结果均吻合良好。本文将纤维取向分布图像处理技术与复合材料力学性能预测方法相结合,有助于更高效准确地理解短纤维增强复合材料的构效关系,对于复合材料结构设计具有较高的指导作用。

     

    Abstract: 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.

     

/

返回文章
返回