An integrated computational framework for predicting and enhancing fiber impregnation in continuous glass-fiber/PLA co-extrusion

The study presents an integrated analytical and numerical framework to predict continuous glass-fiber impregnation in melt polylactic acid (PLA) and validate it with available experimental data. Scanning electron microscope (SEM) image analysis of extruded filaments and dry fiber bundles provided fiber volume fractions and geometry inputs, which were modeled into porous-media parameters. Melt rheology was explored using the Carreau–Yasuda formulation with Arrhenius temperature dependence. Also, the Kozeny–Carman permeability model was evaluated using experimental data. The analytical model mapped the degree of impregnation across a 3 × 3 process window that combines variations of nozzle temperature (463–503 K) and printing speed (100–200 mm/min). Transient ANSYS Fluent simulations performed with the fiber bundle (as a homogenized porous zone) coupled with heat transfer, produced visualizations of melt flow, thermal gradients, and impregnation progression within the hot end. Both analytical and computational fluid dynamics (CFD) implementation models were validated against available experimental data, achieving close agreement with mean absolute errors of 0.0416 and 0.0708, respectively. The impregnation quality improvement with increased nozzle temperature and decreased printing speed due to enhanced polymer mobility and residence time was validated, with error values ranging from 8% to 10% only. The integrated SEM-calibrated analytical and CFD framework provides a predictive tool to refine co-extrusion parameters and minimize voids. The methodology provides guidance on improving hot-end design and 3D printing process parameters for continuous-fiber-reinforced thermoplastic composite.