Assessing the structural performance of additively manufactured carbon fibre reinforced PLA-based adherends bonded with graphene-enhanced adhesive using experimental and ANN analysis

Additive manufacturing (AM) enables the fabrication of complex geometries; however, the mechanical performance of 3D-printed thermoplastic assemblies remains limited by the inherent weakness of 3D-printed polymers and by constraints in joining. The present research addresses these problems by examining the mechanical behaviour of adhesively bonded single-lap joints (SLJs) prepared from carbon-fibre-reinforced polylactic acid (PLA-CF) adherends and epoxy reinforced with graphene nano filler. Dog-bone specimens were fabricated to determine the optimal printing parameters, with a 0° raster orientation showing the highest tensile strength. SLJs with 0.5–1.5 wt.% graphene nano powder (GNP) enhanced epoxy were tested under shear, flexural, and free-vibration loading conditions. The results show that 1.5 wt.% GNP inclusion has the highest improvement, increasing shear and flexural strength by 107.71% and 67.15%, respectively, compared to unmodified joints. Failure analysis indicated that the failure mode shifted from mixed mode to cohesive failure with the addition of 1.5 wt.% GNP, due to microcrack bridging and interfacial debonding. Furthermore, a backpropagation Artificial Neural Network (ANN) model was developed to predict the nonlinear behaviour of the SLJs. The ANN demonstrated excellent agreement with experimental data, with an R2 value of 0.99345. Thereby, the combined experimental and ANN-based approach proves that incorporating short carbon fibres into the adherends and GNP reinforcement into the adhesive can significantly enhance the structural performance of SLJs. The present results will be useful for designing high-strength, lightweight AM-based bonded structures suitable for advanced engineering applications.