Statement of problem
Three-dimensional (3D)–printed casts are used successfully as diagnostic casts in orthodontics. However, whether 3D-printed casts are sufficiently accurate to be used as definitive casts for fixed dental prostheses (FDPs) is unclear.
The purpose of this in vitro study was to evaluate the fit of 3-unit FDPs fabricated on 3D-printed casts made by digital light processing and to investigate the clinical applicability of 3D printing.
Material and methods
A master model was fabricated from epoxy resin. Stone casts were made from dual viscosity impressions (conventional stone cast [CS] group, n=10). The 3D-printed casts were fabricated using a 3D printer after obtaining digital virtual casts by digital scans (3D-printed cast [3DP] group, n=10). All FDPs were fabricated with a 5-axis milling machine. The master model and intaglio surface of the milled FDPs was superimposed using 3D analysis software to measure the accuracy. Two-way ANOVA was performed to identify a significant difference between the groups (3DP and CS) and sides (pontic side, nonpontic side) and their interactive effects (α=.05). The Tukey honestly significant difference test was used for post hoc analysis.
Two-way ANOVA showed significant differences between the 2 groups (3DP and CS) in the marginal and internal root mean square (RMS) values ( P <.001). However, no significant difference was found in the marginal RMS values ( P =.762) between the pontic and nonpontic sides. The 3DP showed significantly higher RMS values than the CS ( P <.001).
The fit of FDPs produced from 3D-printed casts was inferior to that of conventional stone casts; however, all FDPs showed clinically acceptable accuracy. These results suggest that 3D-printed casts have clinical applicability but that further improvement of the 3D printer is necessary for their application in prosthodontics.
Precise casts are essential for fabricating accurate prostheses. If digital scans obtained with an intraoral scanner are routinely used in dental offices, other types of casts would be necessary as alternatives to stone casts. One such alternative is the 3D-printed cast.
The quality of a fixed dental restoration is determined by its marginal and internal fit. Accurate replication is essential to ensure a precise fit. Hence, precision in making impressions and replicas is extremely important.
Definitive casts can be fabricated using the conventional method—elastomeric impression and stone cast—or a digital scan of the teeth with an intraoral scanner. Stone casts have been used for diagnosis, treatment, and prostheses fabrication for many years. However, they are susceptible to damage. Moreover, they are difficult to store as they are bulky. Three-dimensional (3D) digital virtual casts obtained from an intraoral scanner can avoid these problems. They are saved in a digital format and can be transmitted digitally. The data are obtained directly from the teeth, thereby eliminating the need to fabricate a definitive cast after making an impression. However, for some prosthetic procedures, definitive dental casts are still required.
Definitive dental casts can be fabricated from 3D digital virtual casts by subtractive or additive manufacturing. Subtractive manufacturing has simplified the prosthesis fabrication process and improved accuracy. However, reproducing complex shapes and undercuts with this manufacturing method is difficult because the axes of the milling machine are limited. Furthermore, this method results in significant waste and can lead to error (called drill compensation) because of the diameter of the milling bur.
In contrast, additive manufacturing, also known as 3D printing, transforms the designed CAD files into slice data, builds them layer by layer, and creates the desired shapes. Minimal material is wasted, and features such as undercuts and complex internal shapes can be created. Furthermore, several products can be manufactured simultaneously. These advantages have promoted interest in 3D printers, which are being used increasingly in prosthodontics. Moreover, the original patent for 3D printer technology has expired, reducing the costs associated with 3D printers.
Digital light processing (DLP) 3D printing uses the photopolymerization method. The 3D printer flashes a single image of each layer on the entire platform at one time. Therefore, DLP can reduce printing time as all layers are exposed at once. The accuracy of the cast varies with the type of 3D printer. Moreover, the DLP method has been reported to be more accurate than the other types of 3D printers for fabricating diagnosis casts.
At present, 3D printers are mainly used to fabricate orthodontics casts and surgical guides for dental implant surgery. However, they are expected to be used more widely for not only fabricating prostheses but also fabricating definitive casts for prosthodontics.
Studies have evaluated the accuracy of 3D-printed orthodontic diagnostic casts. However, studies on 3D-printed definitive casts for fixed dental prostheses (FDPs) are sparse. Therefore, the purpose of this in vitro study was to evaluate the marginal and internal fit of FDPs fabricated on a 3D-printed cast (3DP) and evaluate the accuracy of 3D-printed casts compared with a conventional stone cast (CS). The null hypothesis was that no statistically significant difference would be found in the marginal and internal fit between groups (CS and 3DP) and sides (pontic side and nonpontic side).
Material and methods
A maxillary resin typodont (AG-3 ZPVK; Frasaco GmbH) was prepared for a 3-unit FDP with the right first premolar and first molar as abutments. The abutment preparations had 1.2-mm, 360-degree chamfer margins. The prepared typodont was duplicated in epoxy resin (Modralit 3K; Dreve Dentamid GmbH) as a master model. To obtain reference data, the model was digitized with an optical scanner (Comet LED 3D scanner; Steinbichler Optotechnik GmbH) with 6-μm precision.
Ten custom trays (SR Ivolen; Ivoclar Vivadent AG) were fabricated after applying a layer of baseplate wax on the master model and coated with tray adhesive (Identium Adhesive; Kettenbach GmbH). Dual viscosity impressions (n=10) of the master model were made according to the manufacturer’s recommendations. To ensure complete polymerization, the impressions were removed from the master model after 10 minutes (3 times longer than the time recommended by the manufacturer), stored for 8 hours at 23 °C, poured with Type IV dental stone (FujiRock; GC), and separated after 45 minutes. After storage for 48 hours, the stone casts were digitized with a reference scanner (Comet LED 3D; Steinbichler Optotechnik GmbH) and saved as a standard tessellation language (STL) file (CS group).
To fabricate the 3D-printed cast (3DP group) (n=10), the master model was scanned by an experienced clinician with an intraoral scanner (CS3500; Carestream Dental LLC). The corresponding exported STL files (digital virtual casts) were used to fabricate 3D-printed casts with a 3D dental model printer (3Dent; EnvisionTEC GmbH) (n=10). The 3D printer had a resolution of 50 μm and printed 10 casts in approximately 1.5 hours. These casts were then digitized with the reference scanner (Comet LED 3D; Steinbichler Optotechnik GmbH) and saved as an STL file (3DP group).
The 3-unit FDPs were designed by CAD software (DentCAD; Delcam PLC) using an STL file (stone cast data and 3D-printed cast data). The occlusal, axial, and margin cement spacers were set as 0 μm, and the radius correction was set at 0.6 mm. The design files were imported with CAM software (GO2dental; GO2cam Intl). The 3-unit FDPs were milled from a polyurethane block (innoBlanc model; innoBlanc GmbH) with the 5-axis milling machine (DWX-50; Roland DG Corp). After milling, the intaglio of the completed 3-unit FDPs was scanned with a reference scanner and saved as an STL file.
For accurate superimposition, all STL files were reduced to the area of interest by removing artifacts and errors using 3D analysis software (Geomagic Verify 2015; Geomagic GmbH). To superimpose data, the master model data and 3-unit FDPs intaglio scan data were first autoaligned. The best-fit alignment command was applied for accurate alignment ( Fig. 1 ). Then, the whole deviation was determined using values from a color map, and the tolerance ranges were set as follows: upper/lower (±10 μm); maximum/minimum (±100 μm). Root mean squares (RMSs) were used to measure dimensional differences between the master model and the intaglio surface data of the digitized 3-unit FDPs. RMSs were calculated by the following equation: