Parametric optimization of material extrusion 3D printing process: an assessment of Box-Behnken vs. full-factorial experimental approach

This work investigates the efficiency of the Box-Behnken design (BBD) in contrast with the full-factorial design (FFD) in ultimate tensile strength (UTS) of PA12 material extrusion 3D printing (ME-3DP) specimens. Three input parameters, i.e., the raster angle (A), layer thickness (B), and nozzle temperature (C) with three levels each, were employed to compare the BBD and FFD efficiency. The 81 full-factorial UTS initial experimental data used in this research have been produced in a previous work published by the authors. Fifteen (15) out of 81 experiments were selected for the BBD design with three repetitions on the central point (0,0,0). Main effect plots (MEP), interaction plots, surface plots, ANOVA analysis, normality plots, mean absolute percentage error (MAPE), and the root-mean-square error (RMSE) evaluate the BBD and FFD approaches. The BBD MAPE and RMSE indexes show that the Box-Behnken design is appropriate for parameter analysis and processing investigation resulting in a 5.3% MAPE and 2.75 RMSE, close to 5.2% and 2.44 of the full-factorial MAPE and RMSE indexes.