The purpose of this research is to optimize the fused deposition modeling (FDM) 3D printing parameters for carbon fiber-reinforced ABS (CF-ABS) composites to significantly enhance their mechanical performance, specifically tensile strength and wear resistance. By leveraging a hybrid particle swarm optimization-artificial neural network (PSO-ANN) model, this study aims to move beyond traditional experimental methods, which are often time-consuming and resource-intensive.
Full factorial design (FFD) methodology is employed to systematically explore the influence of key 3D printing parameters (slice height, infill density and shell thickness) on the mechanical properties of CF-ABS composites. Specimens are fabricated using a Stratasys F170 3D printer, adhering to American Society for Testing and Materials (ASTM) D638 Type IV for tensile strength and ASTM G-99 for wear resistance. Subsequently, experimental data on tensile strength and wear resistance are collected. These data then form the basis for training and validating a metaheuristic hybrid tool, the PSO-ANN. Finally, the optimized PSO-ANN model is used to predict optimal printing parameters for enhanced mechanical properties, with these predictions being experimentally validated.
Experimental results, based on an FFD, showed a maximum tensile strength of 0.038 kN/mm2 at a 0.01-inch slice height, 33.3% infill density and 0.18 mm shell thickness. The minimum wear resistance of 0.005053 mm3/m was achieved at 0.007-inch slice height, 66.6% infill density and 0.12 mm shell thickness. Subsequently, the PSO-ANN model predicted and experimentally validated improved mechanical properties: a highest tensile strength of 0.042 kN/mm2 and a lowest wear rate of 0.004605 mm3/m at optimal parameters of 0.01-inch slice height, 66.6% infill density and 0.06 mm shell thickness.
This research offers significant originality by employing a hybrid metaheuristic approach, specifically the PSO-ANN model, to optimize FDM 3D printing parameters for CF-ABS composites. Unlike conventional trial-and-error or single-factor optimization methods, this integrated computational and experimental framework provides a more efficient and accurate pathway to achieving superior mechanical properties. The study's value lies in its potential to reduce material waste and production time in additive manufacturing, leading to more robust and reliable CF-ABS components for diverse engineering applications, thereby contributing to advancements in material science and smart manufacturing.
Carbon fiber-reinforced ABS (CF-ABS) composites are lightweight and have higher strength and wear resistance, making them suitable for engineering applications.
Optimized 3D printing settings improve material strength and decrease wear.
The study identifies optimal tensile strength for various layer heights, infill density and shell thickness.
A smart optimization technique (particle swarm optimization-artificial neural network (PSO-ANN)) predicts the best settings for strength and durability.
