Nonlinear Elasticity and Damage Prediction in Automated Fiber Placement Composites via Nested Micromechanics

Automated fiber placement (AFP) composites exhibit complex mechanical behaviors due to manufacturing-induced mesostructural variations, including resin-rich regions and tow gaps that significantly influence both local stress distributions and global material responses. This study presents a hierarchically nested modeling framework based on the Parametric High-Fidelity Generalized Method of Cells (PHFGMC) to predict the effective elastic properties and nonlinear mechanical response of AFP composites. The PHFGMC model integrates micro- and meso-scale analyses using representative volume elements (RVEs) derived from micrographs of AFP composite laminates to capture these manufacturing-induced characteristics. Multiple RVE configurations with varied gap patterns are analyzed to quantify the influence of mesostructural features on global stress-strain response. Predictions for linear and nonlinear elastic behaviors are validated against experimental results from carbon fiber/epoxy AFP specimens, demonstrating good quantitative agreement with measured responses. A cohesive extension of the PHFGMC framework further captures damage initiation and crack propagation under transverse tensile loading, revealing failure mechanisms specifically associated with tow gaps and resin-rich areas. By systematically accounting for manufacturing-induced variability through detailed RVE modeling, the nested PHFGMC framework enables the accurate prediction of global mechanical performance and localized behavior, providing a robust computational tool for optimizing AFP composite design in aerospace and other high-performance applications.