A influência da resolução do arquivo .stl na obtenção de modelos digitais e impressos

Abstract

Statemente of the problem. Different .STL file resolutions can directly influence the accuracy of digital models and the achievement of printed models. Objective. The objective is to evaluate the accuracy of digital and printed models obtained from different .STL file resolutions. Material and methods. A single master model was scanned ten times using (CEREC Omnicam, Dendsply Sirona), of these ten scans each was exported in three different resolutions (high, medium and low resolution), totaling 30 resolutions (N=30). For the control group, the master model was scanned with a bench scanner (in Eos X5, Sirona), resulting in only one .STL model (GC). All resolutions were 3D printed (Anycubic Photon S) totaling 30 physical models with resin for dental models (Prizma 3D model, Makertech labs). These models were scanned in a bench scanner to obtain the .STL of the printed models to be evaluated. All file analyzes were performed using the 3D comparison software (Geomatic Control, 3D Systems). To analyze the accuracy of the measurements in each type of resolution, the Shapiro-Wilk normality test was performed. As the ANOVA residuals did not show a normal distribution (p-value <0.05), the non-parametric Kruskal-Wallis test was used to compare the groups to bench scanning, followed by the Dwass-Steel-Critchlow- Fligner for multiple comparisons. To test the reproducibility of the measurements, the t test was used to compare whether the averages of the differences were equal to zero for each type of resolution. For the maximum and minimum means, as the data did not present a normal distribution, the Wilcoxon test was used.

Results. It is verified that for both digital and printed models there was no statistical difference between the groups for all the evaluated resolutions (p-value> 0.05). The average maximum and minimum deviations are close to 50 microns (μm). Conclusions. Different .STL file resolutions had no influence on the accuracy of print and digital models.