Advanced Metaheuristic-Based Optimization for Tensile Strength in Fused Filament Fabricated Acrylonitrile Butadiene Styrene Parts.

Abstract

This study presents a hybrid Artificial Intelligence (AI) strategy to maximize the tensile strength of Acrylonitrile Butadiene Styrene (ABS) parts produced through Fused Deposition Modeling (FDM). Process parameters—such as infill density, extrusion temperature, and layer height—were optimized by integrating Genetic Algorithm-Artificial Neural Network (GA-ANN) modeling, Response Surface Methodology (RSM), and sensitivity analysis. Sensitivity analysis concluded that infill density is the most critical parameter for tensile strength, seconded by layer height and temperature, for steering parameter optimization to better mechanical properties. Optimum parameters (89.99% infill density, temperature 239.99 °C, layer height of 0.266 mm) delivered the highest tensile strength value at 49.35 MPa, marking an improvement by 164% compared to the lowest-performing example (18.7 MPa). Regression analysis validated the GA-ANN model's strength with an R-value of 0.98936. RSM-based ANOVA validated the statistical importance of the model (R² = 0.9987, p < 0.0001). FESEM analysis of tensile fractured test specimens produced using the optimized parameters showed uniform bonding of layers and few voids, supporting the microstructural enhancements responsible for improved tensile properties. This combined optimization approach points to the possibility of substantially enhancing the mechanical properties of FDM-printed ABS parts.