This paper analyses mechanical property prediction through Machine Learning for continuous fiber-reinforced polymer matrix composites printed using the novel Material Extrusion Additive Manufacturing technique. The composite is formed by a nylon-based matrix and continuous fiber (carbon, Kevlar, or fiberglass). From the literature, the elastic modulus and tensile strength were taken along with printing parameters like fiber content, fiber fill type, matrix lattice, matrix fill density, matrix deposition angle, and fiber deposition angle. Such data were fed to several supervised learning algorithms: Ridge Regression, Bayesian Ridge Regression, Lasso Regression, K-Nearest Neighbor Regression, CatBoost Regression, Decision Tree Regression, Random Forest Regression, and Support Vector Regression. The Machine Learning analysis confirmed that fiber content is the most influential parameter in elasticity (E) and strength ($\sigma$). The results show that the K-Nearest Neighbors and CatBoost provided the closest predictions for E and $\sigma$ compared to the other models, and the tree-based model presented the narrowest error distribution. The computational metrics point to a size versus prediction time tradeoff between these two best predictors, and adopting the prediction time as the most relevant criterion leads to the conclusion that the CatBoost model can be considered, when compared to the others tested, the most appropriate solution to work as a predictor in the task at hand.

Fatigue bending tests, under controlled displacement, were performed on a polymer matrix composite material reinforced with continuous Kevlar fibers. The samples were fabricated using the Fused Filament Fabrication (FFF) technique in a Markforged Two^® 3D printer. The static characterization delivered a flexural modulus of elasticity of 4.73 GPa and flexural strength of 110 MPa. The applied loading corresponded to 92.3, 88.5, 86.2, and 84.7% of the static flexural displacement, giving 15, 248, 460, and 711 cycles for failure. Additionally, two numerical models were created: one using orthotropic properties for static loading conditions; and a second one using isotropic in-bulk properties for fatigue modeling. The second model was able to reproduce the experimental fatigue results. Finally, morphological analysis of the fractured surface revealed fiber breakage, fiber tearing, fiber buckling, matrix cracking, and matrix porosity.

La fabricación aditiva (AM),y más específicamente la impresión 3D,ha comenzado una revolución de la industria de la manufactura al proporcionar capacidades de producción para piezas que eran imposibles de fabricarhace algunos años. Una tecnología bastante reciente,desarrollada porMarkforged,ha elevado estas capacidades a un nuevo nivel al permitir la impresión de compuestos de matriz polimérica con refuerzo continuo de fibra. Sin embargo, por ser este un método nuevode fabricación, no existe un modelo consolidado para predecir las características mecánicas ni los modos de falla que presentan al estar sometidas a cargas. El presente trabajo recoge los estudios sobreel daño y falla progresiva en materiales compuestos de fibras largas producidos por manufactura aditiva

This paper shows how temperature influences impact energy for continuous fiber additively manufactured (AM) polymer matrix composites. AM composites were fabricated with a nylon-based matrix and four continuous reinforcements: fiberglass, high-temperature fiberglass (HSHT), Kevlar, and carbon. The tested temperatures ranged from −40 to 90 °C. The chosen printed configuration for the lattice structure and fiber volume was the configuration that was found to perform the best in the literature, with a volumetric fiber content of 24.2%. Impact tests showed that the best response was fiberglass, HSHT, Kevlar, and carbon, in that order. The impact resistance was lowered at temperatures below ambient temperatures and above 50 °C. Additionally, each material’s impact energy was adjusted to third-degree polynomials to model results, with correlation factors above 92%. Finally, the failure analysis showed the damage mechanisms of matrix cracking, delamination in the printing direction, fiber tearing, and fiber pulling as failure mechanisms.