Abstract
To measure and compare the accuracy of 3D-printed materials used for RPD production to improve workflow and eliminate errors in manufacturing.
A partially edentulous maxilla (Kennedy Class III, modification 1) was prepared and designed with proximal plates, rest seats and clasps in one first premolar, one canine and two second molars. A total of 540 3D printed RPD frameworks were 3D printed with three different types of resin (DentaCAST (Asiga, Australia), SuperCAST (Asiga, Australia) and NextDent (3D Systems, Netherlands)). To evaluate the trueness of the printing materials, they were printed with three types of layer thickness: 50 μm, 75 μm and 100 μm, using two types of build angles: 0° and 45° and three types of plate locations: side, middle and corner. After production, all specimens were scanned and superimposed with a control sample that was digitally designed. Using the initial alignment and best-fit alignment method, the root mean square error (RMSE) was calculated. To capture region specific discrepancy, 10 points of XYZ internal discrepancy within RPDs were measured and Euclidean error was calculated. Data was statistically analysed using Shapiro-Wilk and Kruskal-Wallis tests, one-way ANOVA and T-test (SPSS Version 29) and MATLAB (R2022b).
Optimal results were found using 45°, middle of the build plate and layer thicknesses of 100 μm (115 ± 19 μm, DentaCAST), 75 μm (143 ± 14 μm, NextDent), 50 μm (98 ± 35 μm, SuperCAST), which were clinically acceptable. Results were statistically significant when comparing layer thickness in each testing group (p < 0.001). Layer thickness was a primary parameter in the determination of print accuracy among all materials (p < 0.001). Higher discrepancies and failures were observed in 0° prints. No statistically significant difference was found in material usage between build angles or layer thickness (p > 0.005).
All three 3D printing materials exhibited clinically acceptable RMSE results with a build angle of 45° with a printing layer thickness of 50 μm for SuperCAST, 75 μm NextDent and 100 μm for DentaCAST. The highest discrepancies were mostly found in posterior clasps, while the lowest discrepancy was found in palatal straps. Despite unoptimized spacing of prints, frameworks configured to print in the middle of the build plate result in the least printing failures.