Influence of attachment designs in the insertion and removal force of clear aligners: An experimental study of strain sensors

INTRODUCTION

Clear aligners (CAs) provide significant advantages over traditional braces, including enhanced comfort, better esthetics, and easier oral hygiene,^[1,2] leading to their growing popularity among patients.^[3,4] Invisalign® (Align Technology, Santa Clara, CA, USA), established in 1998, has become a leading brand, offering customized treatment plans through advanced software and manufacturing CA made of polymeric material.^[5] Over the years, the adoption of CA has expanded, particularly for treating cases of medium and low complexity.^[6] The CA was named “in-office CA” because they are made in dental offices and allow orthodontists to create their own brand, which enables faster initiation of treatment.^[7]

In-office CA covers the entire tooth surface and part of the gingiva, incorporating grooves for resin-based attachments to facilitate tooth movements and enhance the fit of the aligner.^[5] Successful outcomes depend on proper case selection and meticulous virtual planning.^[8] The design, type, size, and orientation of attachments significantly influence the CA’s performance.^[9] The stiffness and thickness of the polymer used to make the CA also influence the CA’s strain during insertion and removal,^[10,11] directly affecting orthodontic treatment accuracy.^[12]

A systematic review by Iliadi et al.,^[13] found distalization of the upper molar to be the most accurate movement (88% accuracy), while extrusion (30%) and rotation (36%) were less predictable. Despite advancements, 70–80% of aligner patients require refinements or auxiliary treatments, highlighting the challenges posed by polymer limitations, software planning, and patient compliance.^[5,14] Orthodontists are responsible for applying the correct forces to achieve desired outcomes, but the effectiveness of these forces remains contingent on the material properties of CA.^[4] Over time, mechanical, chemical, and physical stressors degrade the polymer’s ability to deliver consistent forces, necessitating precise measurement of force application to ensure treatment efficiency.^[9]

The development of sensor technologies has significantly improved strain measurement capabilities.^[15,16] Recent developments in nanotechnology have enabled real-time quantification of orthodontic forces through multipiezoresistive stress sensors and microchips embedded in aligners.^[17-19] These technologies facilitate precise deformation tracking, providing insights into polymer behavior during CA insertion and removal.^[10,20] This experimental study analyzed the influence of attachment design on the strain and respective expansion of the polymer during insertion and removal of the appliance.

The study hypothesizes that different attachment designs significantly impact the strain, stress, and expansion thresholds of CA during insertion and removal.

MATERIAL AND METHODS

Aligners containing attachments with different designs were manufactured. The aligners were inserted and removed from polymeric models of patients’ arches. The study was approved by the Ethics and Research Committee of the School of Dentistry of the Federal University of Rio de Janeiro (UFRJ), approval number 6.435.151 (CAAE: 71016123.0.0000.0268), necessary because it involved the use of a digital model obtained from a patient’s intraoral scan. The aligners were inserted and removed from polymeric models of patients’ arches. This observational in vitro study adhered to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines^[21] to ensure comprehensive and transparent reporting of the methodology and findings.

The virtual model used met the following inclusion criteria: A healthy adult patient (aged over 18 years) with all permanent teeth present (excluding third molars) and a low complexity rate according to the Index of Complexity, Outcome, and Need.^[22,23] Exclusion criteria included the presence of supernumerary teeth, missing teeth, dental implants, endodontic treatment, restorations made after or during orthodontic treatment, craniofacial deformities, syndromic conditions, and patients under 18 years of age.

An intraoral scan was performed using the Technology Refining Intraoral Outcome Scanning (TRIOS) scanner (3Shape, Copenhagen, Denmark), the upper dental arch was selected, and the data were exported in standard triangle language format for subsequent virtual treatment and model preparation using MeshMixer software (Autodesk Inc., San Rafael, CA, USA) and ArchForm v.2.3.0 (San Jose, CA, USA). Various attachment designs were added to the models [Figure 1 and Table 1]. A total of 10 experimental models were created, featuring attachments on dental elements 11, 12, 13, and 16 of the right hemi-arches, while the left hemi-arch remained unaltered to serve as a control. An additional control model with no attachments was included. The attachments, such as pyramidal and half-moon types, were designed using MeshMixer, while other attachment types were sourced from the ArchForm software library. Each attachment was positioned along the tooth’s vertical axis for consistency. The final sample size of 11 CA was determined based on previous studies employing similar in vitro setups^[2,7,13] to achieve statistical power (80%) while maintaining feasibility.

Close

Figure 1:

Respective position of attachments from each experimental group. Dental unit 11 with (a) Half-moon; (b) Pyramidal; (c) Beveled low profile; (d) Round wedge_2; dental unit 12 with (e) Beveled low profile; (f) Distal slice; dental unit 13 (g) Round wedge_2; (h) Distal slice; and dental unit 16 with (i) Rectangle; and (j) Beveled.

Export to PPT

The digital models were prepared for 3D printing using an Anycubic Photon printer (Anycubic, Shenzhen, China), and then printed horizontally with a 100-micron layer configuration to ensure dimensional accuracy. Anycubic Water-Washable Gray resin was used, and each printing cycle (45 min) produced four models. The printed models underwent cleaning and curing in the Anycubic 2.0 Wash and Cure machine, followed by a thermopolymerization phase using a Thermopressurizer (Flo Laboratory, Brazil) with Polyethylene Terephthalate Ethylene Glycol polymer material of 0.63 mm thickness (Orthomundi, Porto Alegre, Brazil). At the end, the CA were trimmed with goldsmith’s scissors to a 2-mm extended design and sterilized in an ultraviolet chamber.^[9,24]

The CA’s strain during insertion and removal from models was measured using a strain gage. At the Biomaterials Laboratory of the Military Engineering Institute (IME, Rio de Janeiro), unidirectional PA-06-060BA-120-L strain gauges (Micro-Measurements; Vishay Precision Group, Wendell, NC, USA) were glued into CA’s surface. These sensors had a resistivity of 120 Ω, a width of 1.57 mm, a length of 3.81 mm, and an area of 5.982 mm² (Excel Sensores, Embu, Brazil). The strain gauges were firmly affixed using a high-strength, quick-curing cyanoacrylate adhesive, applied to a cleaned and dried bracket surface near the CA’s attachments [Figure 2]. Each gauge was positioned under consistent pressure for 30 seconds to ensure full contact and minimize adhesive thickness. The placement was standardized by using a custom positioning jig, avoiding attachment relief areas to maintain both mechanical integrity and uniform strain transfer. After fixation, the adhesive was allowed to cure for 10 minutes before any measurements were taken, ensuring stability during manual insertion and removal cycles.

Close

Figure 2:

(a) Application of resistant adhesive with paint brush to dental unit 11; (b) Placement of the sensor on the flat surface of the buccal front view of the clear aligner (CA); (c) CA positioned on printed model with strain gauge sensor installed in dental unit 11; and (d) Connection of force strength sensor through copper wire to Spider-8 analysis software.

Export to PPT

The strain gauges were connected through copper wire to the Spider-8 measurement system. The Spider-8 can be used to measure force, strain, pressure, acceleration, and temperature. The Spider-8 was connected to a computer through the printer port. The electronic signals were analyzed using software (v.2.2; Catman Easy, HBM). Calibration was performed before installation, and all measurements followed a standardized protocol. The strain measurements were recorded over six timepoints with 45-s intervals, totaling 5 min of analysis per CA. Measurements included three manual insertions and three removals for each CA. Each CA was inserted and removed 6 times from the model.

The strain gage operates by changing resistance in response to deformation, stretching, or compression. Orthodontic force attenuation was assessed based on the polymer material’s ability to return to its original shape after stabilization during insertion and removal. Deformation (negative values) and expansion (positive values) rates were recorded using Spider-8 software, with the data exported for analysis in Microsoft Excel® 2025. Statistical analyses were performed on all variables. Data normality was verified using the Shapiro-Wilk test, and comparisons were made using repeated measures analysis of variance, with significance set at 0.05 using JAMOVI software (v.2.3).

RESULTS

[Figure 3] presents real-time graphs exported from Spider-8 software, v.2.2 (Catman Easy, HBM). The first measurement in each graph corresponds to the initial insertion of the CA. After 45 seconds, the subsequent measurement captures the first removal. This sequence of insertion and removal was repeated continuously until the 5-minute analysis period was completed for each CA. The [Figure 3 ] displays the recorded variables for deformation stress (negative values) and expansion (positive values) of the polymer material during manual insertion and removal. The elapsed time begins at milestone zero, marking the first insertion of the aligner into the model. The graphs indicate the peak deformation stress during insertion, followed by material stabilization as shown by the dilation (expansion) phase.

Close

Figure 3:

Real-time graph analyses of the deformation stress (−) and expansion (+) of each CA during manual appliance insertion and removal, images retrieved from Spider-8 software. The first row shows CA 1–4 from left to right; the second row shows CA 5–8; and the third row shows CA 9–11. CA: Clear aligner.

Export to PPT

[Table 2] summarizes the average strain during expansion (CA insertion) and contraction (CA removal) close to each attachment type. The control group, without attachments, showed a low mean expansion of 2.730 (Standard deviation [SD] = 0.327) µm/m. The attachments on dental unit 11, designed for extrusion, including beveled low profile, pyramidal, and half-moon types, showed varying results: two designs exhibited deformation stress, while the half-moon type recorded a mean deformation of −13.009 (SD = 2.399).

On dental unit 16, the rectangle attachment had the highest recorded deformation stress (−13.869; SD = 1.198). The same unit’s beveled horizontal scale attachment demonstrated the highest mean expansion (5.337; SD = 0.453), while the beveled low-profile attachment on dental unit 12 recorded a mean expansion of 1.046 (SD = 1.010). For rotational attachments, the distal slice on dental unit 12 exhibited a mean expansion of 0.446 (SD = 1.046). On dental unit 13, the same attachment recorded a mean deformation stress of −1.042 (SD = 0.897).

DISCUSSION

The primary objective of this experimental study was to evaluate the performance of various attachment designs used in thermoformed in-office CA. Attachments were selected based on their common applications in orthodontics and applied exclusively to the right hemi-arch, targeting teeth 11, 12, 13, and 16, while the left hemi-arch served as a control group without attachments.

Extrusion-specific designs, such as rectangular, beveled, pyramidal, and half-moon attachments, were analyzed for their capacity to distribute loads and facilitate controlled movements.^[25] Rotational movements were studied using round wedge_2 and distal slice attachments. Consistent with findings by Hahn et al.^[1] and Iliadi et al.,^[13] the rotational attachments demonstrated lower predictability, as evidenced by the minimal expansion observed in the distal slice attachment on dental unit 12.

Polymeric CA were manufactured with a thickness of 0.63 mm and showed reduced deformation, supporting more consistent force delivery.^[26] As highlighted in prior studies, thinner aligners are more prone to deformation,^[9,24,27] emphasizing the importance of polymer selection. The rectangular attachment on dental unit 16 exhibited the highest deformation resistance (−13.869 µm/m), aligning with previous reports that identified rectangular designs as effective in resisting intrusive and rotational forces.^[7,17]

The beveled attachment on dental unit 16 demonstrated superior expansion properties (5.337 µm/m), underscoring its suitability for extrusion movements. Comparatively, the beveled low-profile attachment on dental unit 12 exhibited reduced expansion (1.046 µm/m), indicating that variations in attachment height and geometry significantly influence force transmission, according to studies that report geometric and positional factors as critical determinants of orthodontic movement efficiency.^[18,23,28]

Central and lateral incisors (dental units 11 and 12) recorded higher deformation, reflecting their susceptibility to irregular force transmission, aligning with reports by Iliadi et al.,^[13] which noted similar vulnerabilities in anterior teeth during orthodontic treatments. Canines (dental unit 13), however, displayed balanced deformation and expansion, emphasizing their biomechanical role as anchorage units. Meanwhile, molars (dental unit 16) exhibited the greatest resistance to deformation, confirming their role in stabilizing aligner movements and distributing forces.^[29]

Rotational movements presented the greatest challenges, with lower predictability compared to extrusion.^[29] For example, the distal slice attachment on dental unit 13 exhibited a deformation stress of −1.042 µm/m, indicating limited efficacy. These findings corroborate earlier reports that rotational movements are less reliable with aligners.^[5,30] The beveled low-profile attachment underperformed compared to standard beveled designs, highlighting the significance of attachment dimensions and positioning in optimizing extrusion movements.

The integration of strain gauge sensors in this study provided real-time data^[30-32] on deformation and expansion, confirming that attachment design plays an essential role in aligner performance. These findings support the hypothesis that the mechanical properties and design of attachments directly affect the aligner’s capacity to transmit forces effectively.^[19] Such methodologies, previously utilized in dental implant and prosthodontic research,^[33,34] were adapted here to account for the unique structure and mechanical properties of CA.

The exclusion of artificial saliva to avoid interference with sensor accuracy is a limitation, as the inclusion of such conditions may produce more clinically relevant data. Future advancements in sensor technology that accommodate wet environments are recommended to enhance the generalizability of the study.^[10,20] In addition, the in vitro nature of the study and the exclusion of other environmental factors (e.g., temperature), and the absence of representation for the periodontal ligament and bone,^[35] suggest that clinical applicability should be further investigated in future studies.

This study contributes valuable perspectives for improving aligner-based orthodontics, particularly for complex cases requiring customized attachment strategies. Future research should focus on evaluating a broader range of aligner materials and thicknesses under varying clinical conditions. Such efforts will refine our understanding of aligner biomechanics and enhance the predictability of orthodontic treatments.^[4,14]

CONCLUSION

CAs exhibit distinct mechanical responses during insertion and removal, with peak deformation stress occurring on initial insertion and stabilization thereafter. Attachments significantly influence this behavior, with certain designs, such as the half-moon and rectangular types, producing higher stress and expansion values. The control group showed minimal strain, confirming the attachments’ role in guiding force distribution. These findings highlight the importance of attachment design in optimizing aligner performance and treatment effectiveness.