This study investigates the turning performance of additive-manufactured polymer-based composite, with particular emphasis on the resulting dimensional error (DE) and surface roughness Ra. Cutting speed, feed rate, and depth of cut were selected as continuous process variables. Subsequently, regression-based modeling was applied to the experimental data, resulting in predictive models with a coefficient of determination (R2) equal to 96.35% and 92.88% for the DE and Ra, respectively. The analysis indicated that depth of cut and cutting speed accounted for more than 86% of the DE model’s explanatory power, while cutting speed, feed and depth of cut contributed approximately 90% to the Ra model. To further evaluate process performance, the Non-Dominated Sorting Genetic Algorithm II (NSGA-II) was employed to determine the Pareto-optimal solutions that simultaneously minimize the dimensional error and the surface roughness. It was found that the optimal solutions are generated with a cutting speed between 120 m/min and 180 m/min, depth of cut below 0.52 mm and feed ranging from 0.05 mm/rev to 0.10 mm/rev. Finally, additional validation experiments confirmed the reliability of the proposed models, yielding mean absolute prediction errors between the measured and estimated values equal to 3% for the dimensional error and 4.8% for the surface roughness.
Tzotzis et al. (Thu,) studied this question.