Revealing the role of Al4C3 in the mechanical behavior of aluminum/graphene composites through machine learning potential-driven atomistic simulations

Abstract

The mechanical behavior of graphene-reinforced aluminum (Al/G) composites is strongly governed by interfacial characteristics, particularly the formation of the Al₄C₃ phase. In this study, a neuroevolution potential (NEP) model was developed to accurately capture the static and dynamic behaviors of Al/G/Al₄C₃ composites, showing excellent agreement with both first-principles calculations and experimental data. Molecular dynamics simulations based on the NEP model reveal that the presence of Al₄C₃ significantly enhances the tensile strength while retaining high ductility under both parallel and perpendicular loading conditions, as well as across various crystallographic orientations at the Al/Al₄C₃ interface. This enhancement is primarily attributed to the formation of strong covalent bonds at the interface, which substantially improve interfacial strength, as confirmed by both tensile and shear loading analyses. Furthermore, the ultimate tensile strength and Young's modulus of the composites are well predicted by the classical rule of mixtures, with load transfer identified as the dominant strengthening mechanism. These findings offer valuable insights into the reinforcing role of the Al₄C₃ phase in carbon-reinforced aluminum composites.