The fatigue test is an essential method used to evaluate the mechanical properties of materials and to assess their fatigue life under cyclic loads. Since the fatigue behavior of 3D-printed parts is highly dependent on fabrication process parameters, this study investigates the effect of infill density percentage and printing direction on fatigue life. Rotary bending fatigue tests were conducted on polylactide dog-bone specimens subjected to alternating peak stresses ranging from 35 to 70 MPa, printed in both horizontal and vertical orientations. The infill densities examined were 25, 50, 75, and 100%. The number of cycles to failure exhibited significant variability due to the 3D fabrication process itself. The S-N curves show that samples printed horizontally have higher strength than those printed vertically. Horizontal specimens generally exhibited greater failure resistance than vertical ones. Even at 25% infill density, horizontal specimens outperformed vertical specimens at 100% infill density, highlighting the influence of orientation. The Wöhler model is used to model the polylactide fatigue behavior. An ANOVA analysis is then performed to assess with higher accuracy the sensitivity of the number of cycles to failure to the applied stress, the infill density, and the printing direction. Regression equations and response surfaces are obtained to enhance the knowledge. ANOVA results revealed R2 values of 76.14% for horizontal printing and 82.75% for vertical printing. For horizontal printing, σ was the most influential parameter (68.44%), while for vertical printing, all terms except f*f were significant, with σ contributing the most (59.19%). Fractography investigation are performed. The cross-sectional views of the broken specimens demonstrated beach marks on the fracture surface of the fibers, indicating cyclic loading. Additionally, cleavage facets observed on the fracture surface confirmed that a brittle fracture had occurred.
