The following 4 materials were tested: 3 resins intended for direct-printed aligners (TC-85DAC [GRA] from Graphy, Seoul, Republic of Korea; OD Clear TF [OC] from 3D Resyns, Barcelona, Spain; and Clear A [CA] from Senertek, Karatas-Izmir, Turkey) and 1 thermoformed material (Duran Plus [DP] from Scheu-Dental, Iserlohn, Germany). All the aligner materials used in this study are presented in Table I .

To obtain a digital model, the maxillary arch of the first author of this study (C.A.D.) was scanned with an intraoral scanner (Trios 3; 3 Shape, Copenhagen, Denmark). For the manufacturing of printed aligners, a design was created using this digital model with a 0.5 mm thickness using orthodontic software (OnyxCeph; Image Instruments, Chemnitz, Germany). Rectangular horizontal attachments were placed on the buccal side of first molars, premolars, and second molars, and a vertical rectangular attachment was placed on the buccal side of the left central incisor ( Fig 1 ).

Then, the aligner design was imported into a software to prepare the file for printing with the chosen printing parameters and adding supports (Chitubox; CBD-Tech, Shenzhen, China). Supports were avoided in essential areas, such as the intaglio of the aligner.

The aligners were printed with Phrozen Sonic Mini 8K (Phrozen Tech Co, Hsinchu City, Taiwan) for GRA, OC, and CA resins. To print the aligners, a 100 μm layer thickness and an orientation of 45° to the build platform were chosen. Twelve copies of the aligner design were printed for each resin in 2 print runs (6 per print run) per resin.

Postprocessing was conducted using the Phrozen washing unit (Phrozen Tech Co) and the Phrozen curing unit (Phrozen Tech Co) for OC and CA aligners. A glass container filled with glycerin for 10 minutes was used during postcuring. These 2 resins did not require the use of a postcuring unit equipped with nitrogen gas. ^, For GRA aligners, according to the manufacturer’s recommendation, Tera Harz Spinner V2 (Graphy) and THC2 curing unit (Graphy) were used. After the curing process, all printed aligners (GRA, CA, and OC) were immersed in boiling water at 100°C for 2 min to standardize postcuring and ensure the removal of any residual unpolymerized resin. For CA and OC aligners, the glycerin used during postcuring was first thoroughly rinsed off under running warm water before this step. After the initial washing, flush cutters were used to remove the supports from each aligner, and a rag wheel provided a final polish only on their cameo.

For the manufacturing of thermoformed aligners, the same digital model obtained from the maxillary arch scan, with the same attachments placed on the same teeth ( Fig 1 ), was printed using model V2 resin, with the model positioned at 0° of angulation to the build platform and a Form 3B printer (Formlabs, Somerville, Mass). After printing, the model was posttreated according to the recommendation of the manufacturer. DP sheets of 0.75 mm were thermoformed on the printed model using a thermoforming machine (Ministar S; Scheu-Dental). The 0.75 mm sheets were selected because this thickness is widely used for clinical clear aligner fabrication and is recommended by the manufacturer for orthodontic applications. After thermoforming, this initial thickness typically results in a final aligner thickness ranging 0.5-0.6 mm over the tooth surface, which corresponds to values reported for commercial systems, such as Invisalign, ClearCorrect, and other commonly used aligners. This final thickness range provides an optimal balance between flexibility and stiffness, ensuring effective force transmission and patient comfort. ^, Therefore, the choice of a 0.75 mm starting sheet was intended to reproduce conditions representative of actual clinical use. The aligners were then cut using a rotating disc and finally polished only on their cameo with a rag wheel. Twelve aligners were prepared.

Final polishing was carried out according to a standardized protocol by a single operator (C.S.) to avoid operator variability.

To facilitate scanning, the aligners of all 4 groups were numbered and individually sprayed with Cerec Optispray (Dentsply Sirona, Charlotte, NC). The spray layer was thinned with an air spray to minimize deviations owing to the spray coating. The spray application was carried out by a single operator (C.S.) to avoid operator variability. Then, to produce standard tessellation language (STL) files, the intaglio of the aligners was scanned using an intraoral scanner (Trios 3; 3 shape). For scanning, 1 aligner was placed on a work surface without any constraints on it, and the camera rotated around the aligner to avoid the risk of deformation. The scanning procedure was carried out by the same operator to ensure repeatability (C.S.).

After this initial scanning, the aligners were cleaned in an ultrasonic bath to remove the spray coating and individually immersed in AS at 37°C and pH of 6.7 for 24 hours, 7 days, and 14 days. The AS was made with the same formula employed by Kotyk and Wiltshire

The aligners of all groups were scanned repeatedly with the same method after the 3 immersion periods to obtain the STL files and evaluate the effect of aging.

To assess the trueness of the aligners, all STL files were imported into morphometric analysis software (CloudCompare, 2.13.2 version; Paris, France). These STL files were individually superimposed on the clear aligner design for the printed aligners and the digital model for the thermoformed aligners.

For the superimposition of printed aligners, first, 7 points were placed on the attachments in the intaglio of the aligner design STL. The same 7 points were placed in the intaglio of the aligner STL obtained after scanning. Then, the software’s algorithm determined the best fit between the 2 files using these 7 points, generating a color map illustrating 3D-surface deviations ( Fig 2 ).

Seven reproducible reference points were selected on the intaglio surface of the aligner design: 6 on the gingival-mesial surfaces of premolar and molar attachments (R0-R2 and R4-R6) and 1 on the gingival-distal surface of the maxillary central incisor attachment (R3). Points were placed in the flat central area of each attachment surface to ensure reproducibility. These landmarks were positioned only once on the baseline (T0) STL file, and the same coordinates were automatically used by the software for subsequent aligner scans (24 hours, 7 days, and 14 days), to avoid variability in landmark placement caused by material deformation over time.

Only the intaglio surfaces of the aligners were scanned and analyzed, as they represent the contact area with the teeth. For thermoformed aligners, the intaglio STL surface normals were inverted in CloudCompare prior to superimposition to match the orientation of the model surface ( Fig 3 ). The entire intaglio surface was included in the analysis without trimming. The 7-point landmark alignment method was employed for all superimpositions, as it ensures reproducible positioning and has been reported to provide consistent accuracy compared with the best-fit methods. ^,

For trueness determination, the root mean square (RMS) value was recorded and compared according to the type of the aligner material and the AS aging duration. This measurement is commonly used in the literature to determine the trueness of an object, as it is not affected by negative or positive values and is considered an absolute error index indicating the magnitude of error. ^,, This index has symmetry, allowing interchangeability of the data.

Here, i: measurement, N: number of measurements, x: reference value, and x ^∧: measured value.

All measurements and superimpositions were performed by a single operator (C.A.D.), an orthodontist with 10 years of clinical and digital orthodontic experience, trained in the use of CloudCompare through certified courses and previous research work.

Intraexaminer reliability was tested by repeating the superimposition by the same examiner (C.A.D.) after an interval of 2 weeks.

The precision of the aligners was also determined. The obtained STL files of each printed or thermoformed aligner were matched to each other within each group (66 combinations per group and aging duration) using the same 7 points employed for the trueness assessment ( Fig 4 ).

For each resin group, 12 aligners were printed in 2 separate print runs (6 per run) under identical printing and postcuring conditions. The precision analysis included all pairwise comparisons between aligners from the same group, regardless of the print run, to assess the repeatability of the printing workflow under standardized conditions.

Then, the RMS value between the 2 aligner scan STL files was noted and compared according to the type of the aligner material and the AS aging duration. Intraexaminer reliability was tested by repeating the superimposition by the same examiner (C.A.D.) after an interval of 2 weeks.

Statistical analysis

Power analysis was conducted to calculate the sample size (n) required to detect a 15% difference between groups, based on the variability reported in a similar study and setting α at 0.05 and power at 80%. By applying the Lehr formula, n = 16/Δ (Δ being the standardized difference δ/σ, in which δ is the target difference and σ is the standard deviation [SD]), the calculated sample size was 12 aligners per group.

The results of trueness and precision are expressed as means and SDs of global millimetric distances and compiled in an Excel spreadsheet (Microsoft, Redmond, Wash). The normal distribution of variables was confirmed by the Shapiro-Wilk test, and equality of variances was assessed by the Levene test. A 2-way analysis of variance (ANOVA), followed by the Tukey honest significant difference test, was also performed to evaluate the influence of the material type and AS aging duration on the trueness and precision of the aligners. In all tests, the significance level was set at P <0.05. Statistical calculations were performed using XLStat (Addinsoft, Paris, France).

The means and SDs of the RMS trueness values are summarized in Table II , with results of the 2-way ANOVA.