Experimental, finite element and machine learning based investigation of adhesive thickness on tensile strength of 3D printed PLA–CF composite lap joints

Abstract

Additive manufacturing enables the fabrication of lightweight composite structures with improved mechanical performance and design flexibility. In fused deposition modelling, carbon fibre reinforced polylactic acid (PLA–CF) has attracted attention because of its higher stiffness and strength compared with conventional thermoplastics. This study investigates the mechanical performance of adhesively bonded PLA–CF single lap joints, with emphasis on the effect of adhesive thickness. PLA–CF specimens were fabricated using controlled FDM parameters and bonded with CY 230-1 epoxy adhesive. Tensile tests were conducted for adhesive thicknesses ranging from 0.2 mm to 1.0 mm to evaluate load-bearing capacity and deformation behaviour. The results showed that adhesive thickness strongly influences joint strength, with the maximum tensile strength of 31.496 MPa obtained at 0.2 mm. Finite element analysis was performed to examine stress distribution and validate displacement behaviour, showing good agreement with the experimental results. Regression and machine learning models, including linear regression, polynomial regression, support vector regression, and random forest regression, were used to predict tensile strength. Random forest regression showed the best predictive performance, demonstrating the value of combining experimental, numerical, and machine learning approaches for optimising additively manufactured composite joints.