Abstract
The objective of this study was to assess the effect of stacking on the dimensional and full-arch accuracy of 3D-printed models, utilising a standardised assessment methodology. A previously validated methodology involving a standard tessellation language image (STL) reference model, comprising seven spheres on a horseshoe base resembling a dental arch, was used. Six 3D-designed STL models were prepared, optimised, and stacked horizontally using 3D Sprint software. The stacking file was transferred to the NextDent 5100 printer to build the physical models. To assess accuracy, a coordinate measuring machine (CMM) measured the diameter of the spheres n=210, and twenty-one vectors extended between the centres of each of the seven spheres (n = 630). When compared to the reference model, significant differences were observed for dimensional (p = 0.006) and full-arch accuracy (p = 0.006) for all stacked models. Additionally, significant differences were observed between the stacked models for the dimensional accuracy between the posterior (p = 0.015), left posterior (p = 0.005) and anteroposterior (p = 0.002). The maximum contraction was observed in the fourth stacked model, which demonstrated the highest median deviation and least precision within the full-arch (MD = 666 μm, IQR = 55 μm), left posterior (MD = 136 μm, IQR = 12 μm), posterior (MD = 177 μm, IQR = 14 μm) and anteroposterior (MD = 179 μm, IQR = 16 μm) arch segments. In general, the anterior and left posterior arch segments recorded the highest contractions with a median deviation of 34 μm and 29 μm, and precision of 32 μm and 22 μm, respectively. Statistically significant differences were observed between the stacked models in terms of dimensional accuracy that were within clinically acceptable thresholds. The greatest contraction was noted in the fourth model, displaying the least full-arch accuracy compared to the other models. Stacked, additively manufactured, full arch models are a viable alternative for diagnostic, orthodontic, and single-unit prosthodontic applications. In contrast, caution should be exercised when utilising stacked models for full arch high accuracy prosthodontic applications. Further research is needed to assess the impact of additional variables including different printers and resins.