Fatigue analysis of 3D-printed polymer materials under various manufacturing parameters utilizing poisson regression method and ANOVA

In this study, we conducted a statistical evaluation of the experimental results from a prior investigation by the authors concerning the fatigue behavior of 3-Dimensional (3D) printed Polylactic Acid (PLA) materials. Stress versus number of cycles (S-N) curves for the sample types were derived using the Poisson regression method. Within the scope of the statistical analysis, we performed average effect plots and an analysis of variance. At the conclusion of the study, we clearly observed the impact of printing parameters on the fatigue of 3D printed PLA materials, especially at low stress amplitude levels. At high stress amplitude values, the effect of printing parameters on the fatigue behavior of 3D printed PLA parts was limited. The analysis revealed that the stress level was the most influential factor determining the number of cycles to failure. While the production parameters of the PLA specimen, such as raster angle and printing speed, significantly impacted the results at a low stress level of 9.13 MPa, their effect was much less pronounced at a higher stress amplitude of 18.25 MPa, where variations in the number of cycles to failure among different specimen types were minimal. The analysis determined that the optimal parameter combination for 3D printed PLA fatigue specimens was a printing speed of 20 mm/s, a raster angle of 30°, and a stress amplitude of 9.13 MPa. According to the analysis of variance results, the parameter that most significantly influenced fatigue life was the stress level, contributing 54.93%. Raster angle and printing speed contributed 14.52% and 4.19%, respectively. The Poisson regression method proved to be an effective tool for plotting S-N graphs.