3D-printing versus milling for zirconia restorations: artificial aging effect on the mechanical behavior of the restorations bonded to the dentin analogue

This study aimed to compare the mechanical behavior of zirconia restorations produced by 3D printing and milling luted on a substrate made of fiberglass reinforced resin epoxy and subjected to different storage conditions. For this, 3 mol% yttria-stabilized tetragonal zirconia polycrystal (3Y-TZP) samples (N = 64) were produced either by milling (IPS e.max ZirCAD MO) or by 3D printing (LithaCon 3Y 210), with dimensions of 10 mm in diameter and 1 mm in thickness. The specimens were allocated into four groups according to the factors “manufacturing methods” and “storage condition”: ML (milling), ML-AG (milling with aging of the bonded samples), 3D (3D printing), and 3D-AG (3D printing with aging). Discs made of fiberglass reinforced resin epoxy (Ø = 10 mm; thickness = 2 mm) were obtained and adhesively bonded to the zirconia restorations using a resin cement system. Subsequently, half of the specimens were assigned to baseline conditions (stored in distilled water at 37 °C for 24 h), while the other half were subjected to aging conditions (stored in distilled water at 37 °C for 65 days followed by 25,000 thermal cycles between 5 °C and 55 °C). The luted assemblies (n = 15) were then subjected to a cyclic fatigue test using a step-stress methodology (frequency: 20 Hz; initial load: 400 N for 10,000 cycles; step increments of 100 N up to 700 N, followed by 50 N increments until failure). Fatigue failure load (FFL) and cycles to failure (CFF) were recorded for statistical analysis. Surface roughness (Ra and Rz) was evaluated, and topography and fractographic analyses were qualitatively assessed. Data were analyzed using the Mann–Whitney test for roughness, whereas fatigue outcomes (FFL and CFFL) were analyzed using Kaplan–Meier survival analysis and Weibull statistics (α = 0.05). The cyclic fatigue test results demonstrated that both manufacturing methods and storage conditions significantly influenced the mechanical behavior. At baseline, the 3D group exhibited higher FFL and CFF values compared to the ML group, indicating superior mechanical behavior. However, greater variability was observed for the 3D group, as reflected by its lower Weibull modulus. After aging, both ML-AG and 3D-AG groups showed significantly higher FFL and CFF values, with no statistical difference between them. Weibull modulus analysis revealed that the 3D-AG group presented the highest reliability, with no overlap between the confidence intervals compared to the other groups. Surface roughness analysis showed that ML exhibited higher Ra and Rz values than 3D (p < 0.001), which was consistent with the topographic findings, where ML presented more surface irregularities, while 3D showed a more homogeneous surface. Fractographic analysis indicated that failures predominantly originated at the cementation interface. Thus, the manufacturing method influenced the mechanical behavior of zirconia restorations, with 3D printing demonstrating superior performance at baseline, although with greater variability. Aging improved the mechanical behavior and reliability of both groups, resulting in similar behavior regardless of the manufacturing method.