General new geometrical modelling for composite materials: application for knitted composites

Abstract This work introduces a novel geometrical and analytical modeling framework for predicting the stiffness properties of knitted composites, a class of materials known for their complex architectures and versatile mechanical behavior. The proposed approach integrates multi-scale modeling techniques with finite element homogenization to derive an accurate stiffness matrix that accounts for yarn geometry, stitch configuration, and material properties. A user-friendly “click-and-drop” interface was developed to streamline the design and construction of knitted composite geometries, incorporating automated validation checks to ensure structural accuracy. Analytical formulations were combined with discretization techniques to evaluate layer-by-layer contributions and compute macroscopic stiffness with high precision. The model was validated against experimental benchmarks from Huang, Gommers, and Ramakrishna for plain knitted, weft-knitted, and glass-fiber-reinforced composites, yielding prediction errors consistently below 5 % for longitudinal, transverse, and shear moduli. These results confirm the robustness of the proposed framework and its ability to replicate real-world mechanical behavior. The methodology offers a significant improvement over traditional modeling approaches by providing greater accuracy, adaptability to various composite configurations, and reduced computational complexity. This research contributes a generalized, scalable, and practical tool for engineers and researchers, enabling efficient analysis and optimization of knitted composites for applications in aerospace, automotive, biomedical, and wearable technologies. Future developments may extend the model to account for nonlinear, viscoelastic, and failure behaviors, further enhancing its utility for advanced material design.