Biomechanical analysis of retracting maxillary incisors using clear aligners with palatal mini-implants in extraction cases: a finite-element analysis Authors: Zixuan Zhao, Zhiwei Wang, Lei Liang, Lu Liu, Yuan Cao, Fang Jin Published: 2026 ABSTRACT BackgroundAt present, the clinical orthodontic effects associated with the application of clear aligners to extraction cases featuring retraction of the anterior teeth remain a cause for concern. In the present study, the biomechanical effects of this strategy on the maxillary dentition, periodontal ligament, and clear aligners were analysed when palatal micro-implants were used in combination with different traction methods.MethodsSeveral models, including the maxillary dentition (without first premolars), maxilla, periodontal ligaments (PDLs), and aligners, were constructed and imported to finite element software. Four model groups were created: (1) G1 (pure aligner retraction force), (2) G2 (the combination of aligner retraction force with elastic force applied between the palatal mini-implant and the traction button), (3) G3 (the combination of aligner retraction force with elastic force applied between the palatal mini-implant and the precision cutting), and (4) G4 (the combination of aligner retraction force with passive ligation, i.e., rigid wire ligation between the mini-implant and the lingual button on the second premolar). The magnitude of the elastic traction force was increased from 100 g to 250 g, with a gradient increment of 50 g. Subsequently, finite element analysis was performed to simulate the biomechanical effects.ResultsCompared with G1, the incisors in G2 exhibited a larger incisor retraction distance and better vertical control. Although incisor retraction in G3 was slightly lower than that in G2, this method demonstrated better control of anterior teeth torque and anchorage of the posterior teeth. In G4, there was a tendency for a slight increase in incisor retraction compared to G1, while vertical extrusion and lingual tipping of the incisors were the most severe. In G2 and G3, the magnitude of the palatal elastic retraction force within the range of 100 to 150 g was conducive to the retraction of incisors, torque and vertical control of anterior teeth.ConclusionsThe combination of aligner retraction force with an appropriate retraction force applied to palatal micro-implants was beneficial for improving the efficiency of incisor retraction, as well as enhancing both vertical control and torque control. However, appropriate overtreatment should be designed to ensure bodily retraction and reduce tipping movement. Background Clear aligner therapy (CAT) has gradually gained popularity in orthodontic treatment due to its outstanding aesthetic appearance, patient comfort, and ease of maintaining oral hygiene [1]. However, retraction of the anterior teeth to improve a patient’s aesthetic appeal is considered arduous in premolar extraction cases, with frequently reported difficulties, including “the pendulum effect” and “the roller-coaster effect” [1–3]. Given the application of an aligner to the crown of the teeth and a lack of stiffness in clear aligner appliances, the lack of root control may cause a tipping movement of the anterior teeth, torque loss, and deepening of the overbite during the application of retraction forces [2, 4]. Compared to CAT, fixed appliances exhibit greater anterior torque control and improved stability of the posterior anchorage teeth [5]. Compared with the mandibular dentition in premolar extraction cases, Liu et al. reported that the maxillary dentition presented with less anterior teeth retraction and greater posterior teeth mesialization during anterior retraction without a micro-implant [6]. Some clinical studies have reported that to achieve bodily anterior teeth retraction and stronger posterior teeth anchorage, it is necessary to combine protocols, such as staged retraction. Setting an appropriate gap between anterior teeth via sequential distal movement can increase the wrapping area of the aligner, thereby enhancing the efficiency of anterior teeth retraction and facilitating the control of anterior tooth torque and vertical position [2, 3]. Cheng et al. [7] found that according to the angle of ∠U1-SN, adjusting the height of the power ridge in the inner surface of the aligner can achieve the bodily movement of incisors during retraction, while increasing the risk of mesial tipping tendency of molars. Appropriate aligner overtreatment (e.g., 4.8°/mm in this model) with canine attachments was found to alleviate the roller-coaster effect and facilitate the bodily retraction of anterior teeth [8, 9]. Wang et al. [9] found that simultaneous buccal and palatal attachments with an overtreatment of 2.4° on second premolars might be helpful in increasing aligner wrapping and reduce unwanted mesial tipping of posterior teeth during anterior teeth retraction. Nevertheless, these disadvantages associated with the use of clear aligners to close an extraction space still remain challenging for orthodontists in clinical practice. Compared with conventional methods, orthodontic miniscrew implants (OMIs) represent a convenient, simple, and effective method for anchorage control during anterior teeth retraction, with superior performance [10–12]. The application of OMIs for maximum anchorage can reduce anchorage loss by reducing mesial tipping of the posterior teeth and by enhancing anterior retraction. Furthermore, in some cases, OMIs can reduce therapy time [13]. In the hard palate area of adults, the bone thickness is sufficient, the bone density is dense, the keratinized mucosa is tough and thin, without a submucosal layer, and it is far away from important blood vessels and nerves [14]. Therefore, the probability of post-implantation inflammation is very low, making it an ideal area for implant placement. Overall, the failure rate of mini-implants implanted in the palate is 4.7% [12]. A recent systematic review reported that the overall failure rate of OMIs when inserted into the palate was lower than that at maxillary buccal insertion sites. Compared with buccal miniscrews, the combination of a clear aligner with palatal miniscrews and precision cuts is more conducive to improving the efficiency of maxillary molar distalization [15]. Despite the valuable insights gained from previous studies, the specific biomechanical mechanisms associated with the use of lingual buttons and the combined use of precision cuts with OMIs inserted in the palate during maxillary anterior retraction with clear aligners have yet to be fully elucidated. Of the various analytical tools that can be used to investigate orthodontic tooth movement, along with the subsequent stress and strain imposed on the periodontal ligament (PDL) and clear aligner (CA), finite element analysis (FEA) appears to be the most common [16, 17]. In this study, we performed FEA to provide new solutions and biomechanical evidence for CAT in extraction cases that are prone to the roller-coaster effect and pendulum effect. Materials and methods An adult orthodontic patient with a Class I skeletal and dental relationship, requiring the extraction of four first premolars, was identified in the patient archive of the Stomatology Hospital of the Fourth Military Medical University. This patient had permanent dentition, normal teeth morphology, and no periodontal disease. This study was approved by the Ethics Committee of the Stomatology Hospital of the Fourth Military Medical University. First, a cone-beam computed tomography (CBCT) scan was performed. The raw volumetric DICOM data from the scan were imported into Mimics version 21.0 software (Materialize Software, Leuven, Belgium) to create a three-dimensional (3D) geometric maxillary model. The soft tissues, cortical bone, cancellous bone, and teeth were isolated in Mimics software via threshold-based segmentation based on differences in grayscale. Then, the soft tissues were removed. An integrated 3D geometric surface model of the maxilla and dentition was then imported into Geomagic Studio 2015 software (3D System, USA) to further improve the precision of dentition modeling and to construct a standard first premolar extraction model. The first premolars were removed to generate the first premolar extraction model. The canines were moved toward the extraction spaces to create a 1 mm gap between the distal surfaces of the canines and the mesial surfaces of the lateral incisors. The PDL was then modeled on the outer surface of the root with a uniform thickness of 0.25 mm in accordance with a previous study [9]. On the periphery of the cancellous bone, the thickness of the cortical bone shell was set at 1.5 mm. Finally, a high-quality maxillary dentition model (referred to as dentition I) was constructed, including the maxilla (Fig. 1A), dentition (Fig. 1B), and periodontal ligament (Fig. 1C). Fig. 1Models for FEA. A The maxillary FEA model were constructed on the basis of patient data, and the CA, miniscrews, and traction buttons were constructed according to available products. B 3D Models for FEA in different views Models for FEA. A The maxillary FEA model were constructed on the basis of patient data, and the CA, miniscrews, and traction buttons were constructed according to available products. B 3D Models for FEA in different views Next, a new extraction dentition model (referred to as dentition II) was created by retracting the incisors 0.25 mm in the sagittal direction. CAs were modeled on the external surfaces of the tooth crowns with a thickness of 0.5 mm by application of Geomagic Studio 2015 software [15]. Considering the anatomical safety of the palatal structures [18] and the clinical practicability of closing the extraction spaces, the palatal interradicular space 8 mm away from the crest of the alveolar ridge between the first molar and the second molar was selected as the insertion site [19]. The miniscrew was developed according to the Vector TAS orthodontic implant (Ormco, Glendora, Calif) with a total length of 10.7 mm, a diameter of 1.4 mm, and implantation depth of 8 mm (Fig. 1E). NX 1911 (Siemens PLM Software, Germany) was then used to assemble the constructed components into research models, including traction buttons (a generic term for resin buttons bonded to aligners to facilitate elastic traction from palatal mini-implants) (Fig. 1F), precision cuts and miniscrews (Fig. 1G-I). The clear aligner activation protocol was as follows: (1) Digital activation: incisors were retracted 0.25 mm along the Y-axis as a target dentition II to generate a 0.5 mm-thick aligner in Geomagic Studio (Fig. 2A). A fixed boundary condition was set at the base of the maxilla to prevent the model from being influenced by body motion while force was loaded (Fig. 2B). A schematic diagram of the four groups is shown in Fig. 2C. The 0.25 mm positional discrepancy induced elastic deformation of the aligner and the traction force exerted by the mini-implants, creating initial stress fields. (3) FEA calculation: The teeth-PDL-bone system response was calculated in the initial stress fields. This simulated the clinical “activation” of thermoformed aligners on teeth. Model structures were imported into Ansys workbench 2019 software (ANSYS, USA) to generate a 3D finite element model by a meshing process and discretization. The models were then meshed into tetrahedral quadratic elements that were suitable for irregular geometries and the finite element simulation of contact problems. Property parameters utilized in previous orthodontic FEA studies were applied [2, 17, 20]. The discretization parameters, including the number of nodes and elements, are listed in Tables 1 and 2. Fig. 2Retraction protocols. A Incisors were retracted by 0.25 mm along the Y-axis as dentition II to generate a 0.5 mm-thick aligner in Geomagic Studio. B Schematic diagram of boundary condition settings for FEA. C Schematic diagram showing the different groups according to different anterior teeth retraction methods. D The three-dimensional coordinate systems responsible for movement of the maxillary teeth Retraction protocols. A Incisors were retracted by 0.25 mm along the Y-axis as dentition II to generate a 0.5 mm-thick aligner in Geomagic Studio. B Schematic diagram of boundary condition settings for FEA. C Schematic diagram showing the different groups according to different anterior teeth retraction methods. D The three-dimensional coordinate systems responsible for movement of the maxillary teeth Table 1Material properties of the finite element modelComponentYoung’s modulus (MPa)Poisson’s ratioCortical bone13,7000.26Cancellous bone13700.3Teeth19,6000.3Traction button11,4000.35Periodontal ligament0.680.48Clear aligner8160.36Miniscrew114,0000.34Passive metal ligation200,0000.3 Table 1Material properties of the finite element modelComponentYoung’s modulus (MPa)Poisson’s ratioCortical bone13,7000.26Cancellous bone13700.3Teeth19,6000.3Traction button11,4000.35Periodontal ligament0.680.48Clear aligner8160.36Miniscrew114,0000.34Passive metal ligation200,0000.3 Material properties of the finite element model Table 2Number of nodes and elements of the components used in the finite element modelComponentElement no.Node no.G1G2G3G4G1G2G3G4Cortical bone14,61316,19816,20216,19228,08530,75430,76030,761Cancellous bone16,71919,50619,24819,73929,19233,65133,28433,992Teeth50,99550,99550,99550,99581,01081,01081,01081,010Traction button04799058550732608879Periodontal ligament61,95861,95861,95861,958124,671124,671124,671124,671Clear aligner17,54117,54117,99317,55934,92134,92127,91834,984Miniscrew18,03718,03717,55017,96228,00428,00435,02427,884Passive metal ligation00018200009543 Table 2Number of nodes and elements of the components used in the finite element modelComponentElement no.Node no.G1G2G3G4G1G2G3G4Cortical bone14,61316,19816,20216,19228,08530,75430,76030,761Cancellous bone16,71919,50619,24819,73929,19233,65133,28433,992Teeth50,99550,99550,99550,99581,01081,01081,01081,010Traction button04799058550732608879Periodontal ligament61,95861,95861,95861,958124,671124,671124,671124,671Clear aligner17,54117,54117,99317,55934,92134,92127,91834,984Miniscrew18,03718,03717,55017,96228,00428,00435,02427,884Passive metal ligation00018200009543 Number of nodes and elements of the components used in the finite element model A mesh convergence analysis was performed to evaluate the sensitivity of the model to mesh density and to ensure the reliability of the finite element results. During the analysis, local mesh refinement was mainly applied to the periodontal ligament (PDL), with element sizes of 0.8 mm, 0.5 mm, and 0.2 mm, while the mesh settings for the remaining structures were kept unchanged (Table 3). Under identical loading conditions, the displacement of the maxillary central incisor and the maximum root stress were extracted as evaluation indicators for comparison among different mesh configurations. The results demonstrated that both displacement and stress values gradually converged as the PDL mesh was refined. Compared with the 0.8 mm mesh, the total number of elements in the 0.2 mm mesh increased by approximately 294%, whereas the maximum root stress increased by only 12.6%. In addition, the displacement of the maxillary central incisor changed by only 1.6%, indicating that further mesh refinement had a limited influence on the computational results and that the model had achieved satisfactory convergence. Considering both computational accuracy and efficiency, the final mesh configuration adopted in this study consisted of a 2 mm element size for the alveolar bone, 0.5 mm for the periodontal ligament, and 1 mm for the teeth and aligner. Table 3Mesh parameter settings of the finite element modelGroupElement size (mm)Mesh no.Maximum displacement of central incisor(mm)Maximum root stress of central incisor(MPa)Alveolar bonePDLCAToothNode no.Element no.Mesh A20.811227,474129,6110.141790.11928Mesh B20.511292,033161,8260.139510.23660Mesh C20.2111,152,752678,4070.099420.26641 Table 3Mesh parameter settings of the finite element modelGroupElement size (mm)Mesh no.Maximum displacement of central incisor(mm)Maximum root stress of central incisor(MPa)Alveolar bonePDLCAToothNode no.Element no.Mesh A20.811227,474129,6110.141790.11928Mesh B20.511292,033161,8260.139510.23660Mesh C20.2111,152,752678,4070.099420.26641 Mesh parameter settings of the finite element model The material properties of the models were assumed to be homogeneous, isotropic, linearly elastic, and material constitutive [21]. In addition, a friction coefficient contact condition (µ) of 0.2 was set between the CA and the dentition [7]. After configuring the loading parameters, the finite element models were calculated using Ansys workbench 2019 software to generate a series of nodal displacements and von Mises stress distributions. The right maxillary dentition, PDL, and CA were taken as the research objects. As displayed in Fig. 2C, two types of coordinates were designed. First, a global coordinate system (XYZ) was established: the Y-axis was the intersection of the sagittal plane and the occlusal plane, with the positive direction pointing posteriorly; the X-axis, coronally perpendicular to the Y-axis, pointed to the patient’s left in the positive direction; the Z-axis, vertically perpendicular to the X-axis and Y-axis, pointed apically in the positive direction. Second, local coordinate systems were established at the midpoint of the incisal edge and root apex for anterior teeth, and at the center of the occlusal surface and root apices for multi-rooted posterior teeth, to calculate individual tooth torque changes: x (mesiodistal direction), y (labiolingual direction), and z (tooth long axis). Torque changes were calculated based on the angular displacement around the tooth long axis, derived from the difference in sagittal displacement between the crown and root. The mesial point, midpoint, and distal point of the incisal and the apical point of the root of the anterior teeth, the center point of the occlusal surface and multiple apical points of posterior teeth were used as measuring points. Results Differences in the displacement tendencies of teeth during anterior retraction with different traction methods FEA results for tooth displacement in the four groups are presented in Fig. 3. Displacement of the incisors exhibited more intuitive differences (Fig. 3A). In G2 and G3, central incisors showed less retraction but more extrusion and lingual tipping compared to G1, whereas lateral incisors exhibited greater retraction with improved vertical and labiolingual control. In G4, the central incisors exhibited greater retraction and better vertical control, while lateral incisors more obvious extrusion and lingual tipping. The canines exhibited smaller extrusion and better labiolingual control in G2 and G3, while the canine showed the most obvious extrusion in G4. As shown in Fig. 3A, the posterior teeth movement of G4 showed smaller molar mesial movement and effective anchorage protection, while there was no significant difference in the displacement direction of the posterior teeth among the other three groups. Fig. 3Tendency for displacement of the maxillary teeth in the four groups. A The displacement tendencies of the maxillary teeth in each group. B Average displacement of the maxillary teeth in the sagittal direction [(Crown + Root) / 2]. C Average displacement of the maxillary teeth in the vertical direction [(Crown + Root) / 2]. D Degree of the incisors extrusion (vertical average movement / sagittal average movement of incisors). E The average value of torque angulation of central incisors and lateral incisors Tendency for displacement of the maxillary teeth in the four groups. A The displacement tendencies of the maxillary teeth in each group. B Average displacement of the maxillary teeth in the sagittal direction [(Crown + Root) / 2]. C Average displacement of the maxillary teeth in the vertical direction [(Crown + Root) / 2]. D Degree of the incisors extrusion (vertical average movement / sagittal average movement of incisors). E The average value of torque angulation of central incisors and lateral incisors Figure 3B and C show that, in contrast to G1, the central incisors in G4 exhibited more retraction and intrusion, along with the minimal retraction and greater extrusion of the lateral incisors. Lateral incisors exhibited the greatest retraction and significant intrusion in G2 and G3. Canines and posterior teeth exhibited minimal sagittal movement in G4. Canines and the second premolar exhibited reduced vertical movement in G3, while the first molar and second molar exhibited minimal vertical movement in G4 (Fig. 3C). In general, as shown in Fig. 3D, under the same retraction distances, the tendency for vertical movement in the anterior dentition was lowest in G2. As shown in Fig. 3E, minimal changes in the torque of incisors were observed in G3. Appropriate elastic force was beneficial for the control of vertical direction and torque in the anterior teeth and the control of posterior teeth anchorage For incisors under equal elastic force (Fig. 4A), sagittal retraction in G2 was larger than that for G3 at 100–250 g, and the sagittal retraction gradually increased with the rising gradient of elastic force. In terms of vertical movement, the extrusion of incisors under the equal elastic force in G2 was less than that in G3 between 100 g and 250 g. In addition, an increase in the elastic force resulted in a gradual increase in lingual tipping of the incisors. Under equal force, the change in torque angle in G3 was less than that in G2 (Fig. 4C). Fig. 4Tendency for displacement of the maxillary teeth in different groups with increasing elastic force. A The sagittal movement of incisors [(central incisor + lateral incisor) / 2]. B Vertical movement of the incisors. C Changes in torque of the incisors [(central incisor + lateral incisor) / 2]. D Sagittal movement of the canine. E Vertical movement of the canine. F Changes in torque angulation of the canine. G Sagittal movement of the posterior teeth [(second premolar + first molar + second molar) / 3]. H Vertical movement of the posterior teeth [(second premolar + first molar + second molar) /3]. I Changes in torque angulation of the posterior teeth [(second premolar + first molar + second molar) / 3] Tendency for displacement of the maxillary teeth in different groups with increasing elastic force. A The sagittal movement of incisors [(central incisor + lateral incisor) / 2]. B Vertical movement of the incisors. C Changes in torque of the incisors [(central incisor + lateral incisor) / 2]. D Sagittal movement of the canine. E Vertical movement of the canine. F Changes in torque angulation of the canine. G Sagittal movement of the posterior teeth [(second premolar + first molar + second molar) / 3]. H Vertical movement of the posterior teeth [(second premolar + first molar + second molar) /3]. I Changes in torque angulation of the posterior teeth [(second premolar + first molar + second molar) / 3] For the canines, an increase in the elastic force resulted in a gradual strengthening of sagittal control and a reduction in mesial displacement (Fig. 4D). Minimal vertical displacement of the canine was observed in G2 (250 g) and G3 (150 g) (Fig. 4E). The changes in torque shifted from buccal tipping towards lingual tipping as the elastic force was increased; the smallest value was observed at 150 g in both G2 and G3 (Fig. 4F). With regard to the posterior teeth in G2 and G3, as the elastic force increased, the sagittal, vertical and torque control gradually strengthened. Under equal force, the anchorage control in group G3 was superior to that in group G2 (Fig. 4G-I). Appropriate elastic force facilitated vertical control of the incisors during anterior teeth retraction, while excessive elastic force increased the lingual tipping tendency of the incisors As shown in Fig. 5A, the vertical extrusion tendency of incisors in G2 and G3 was reduced under the same retraction distances with an increase in the elastic force. The minimal value was observed in G2 (250 g) and G3 (250 g) (Fig. 5A). Meanwhile, within the force range (100–250 g), the degree of incisor extrusion in G2 and G3 was lower than that in G1 and G4, thus indicating that vertical control was superior to that of G1 and G4. In addition, an increase in elastic force resulted in a gradual increase in the lingual tipping of incisors in G2 and G3. Compared with G1, when the elastic force ranged from 100 g to 150 g, the average torque control of the incisors in both G2 and G3 was better than that of G1. Under equal force, the change in torque angle in G3 was less than that in G2. However, when the traction force ranged from 150 g to 250 g, the lingual tipping of incisors in G2 and G3 further increased, with a worsening tendency of torque control. Fig. 5A The degree of extrusion under conditions of similar sagittal movement (the degree of extrusion = vertical movement / sagittal movement). B Changes in torque of the incisors [(central incisor + lateral incisor) / 2] A The degree of extrusion under conditions of similar sagittal movement (the degree of extrusion = vertical movement / sagittal movement). B Changes in torque of the incisors [(central incisor + lateral incisor) / 2] As the gradient of elastic traction force increased, the maximum principal stress on the PDL gradually increased. With regard to maximum principal stress in the CA, stress in the precision-cutting area of G3 increased drastically with increasing elastic force As shown in Fig. 6A, the distribution of tensile stress was primarily localized on the cervical third of the labial surface of the lateral incisor in G1. The distribution of periodontal tensile stress on the lateral incisor in G2 and G3 was similar to that of G1; however, tensile stress was concentrated on the cervical third of the labial surface of the central incisors in G2 and G3. Compressive stress was predominantly observed at the cervical third of the lingual surface of the lateral incisor. In addition, the distribution of periodontal stress on the canine and posterior teeth was more uniform in G2 and G3 than in the other groups. The maximum compressive stress on the PDL was the smallest in G1. In G2 and G3, the maximum compressive stress on the PDL increased with increasing elastic force (Fig. 6C). Fig. 6The distribution of maximum principal stress in the different experimental models. A The distribution of maximum principal stress of the PDLs across the whole dentition. B The distribution of maximum principal stress on the CA. C The maximum values of maximum principal stress on the PDLs. D Average values of maximum principal stress on the CA The distribution of maximum principal stress in the different experimental models. A The distribution of maximum principal stress of the PDLs across the whole dentition. B The distribution of maximum principal stress on the CA. C The maximum values of maximum principal stress on the PDLs. D Average values of maximum principal stress on the CA The distribution of maximum principal stress on the CA in G1 and G4 was similar, and the distribution of stress was relatively uniform across the entire aligner. The maximum principal stress on the CA was predominantly distributed in the anterior teeth and premolar areas in G2, but was predominantly distributed in the precision cut area in G3 (Fig. 6B). The maximum stress value on the CA was the smallest in group G1, followed by group G4. As the elastic force gradually increased, the magnitude of this stress on the CA also increased in groups G2 and G3. In particular, the magnitude of stress in G3 was significantly greater than that in the other three groups (Fig. 6D).、. Discussion Clear aligners have been increasingly applied in mild malocclusion cases due to their aesthetics, comfort, fewer clinical appointments, and shorter overall treatment time [22, 23]; however, the efficacy of this clinical treatment has raised concerns due to varying predictability for tooth movement in CATs [6]. Especially for some complex extraction cases, compared with conventional fixed appliances, CATs have been associated with more severe pendulum effect and roller-coaster effect, requiring more time for the closure of extraction spaces and the adjustment of torque and occlusion contacts [24]. The core challenge is how to improve the efficiency of anterior teeth retraction while reducing the tendency for anterior teeth extrusion and lingual tipping [25], and enhancing posterior teeth anchorage [17]. However, it is difficult for CATs to control extrusion and buccolingual inclination of the anterior teeth [26]. A previous study revealed that strong direct and indirect buccal anchorages were conducive to enhancing anterior teeth retraction, reducing anterior extrusion and protecting anchorage of the posterior teeth. In addition, strong buccal anchorages could increase the lingual tipping tendency of the anterior teeth and buccal tipping tendency of the posterior teeth [17]. However, anchorage generated by palatal miniscrews with elastic retraction force appeared to mitigate these side effects. As shown in Fig. 3, a palatal elastic retraction force of 100 g enhanced torque control as well as the vertical control of incisors in both G2 and G3. The passive metal ligation exhibited the least favorable vertical control for incisors among the four groups, although this method also contributed to incisor retraction (Fig. 3). Therefore, retracting anterior teeth with CAs, loading a mild elastic retraction force loaded by a palatal anchorage screw, using either precision cut or a traction button, is more conducive to the vertical and torque control of incisors. In general, our results indicated that the traction button favored anterior retraction and vertical control, and was more suitable for extraction cases with anterior teeth protrusion and deep occlusion, requiring simultaneous retraction and intrusion. Given its poor torque control, more anterior overtreatment is required to resist lingual tipping. A study suggested that with vertical rectangular canine attachments, an overtreatment of 4.8° per millimeter of incisor retraction should be designed to roughly ensure bodily retraction [8]. In our study, the precision cut method was superior for incisor torque control and enhancing posterior anchorage. This method might be better suited for the retraction of anterior teeth that are upright or lingually inclined. In comparison to the traction button, however, its application could potentially prolong orthodontic treatment time. In this study, we aimed to further identify the appropriate magnitude of elastic retraction force for different methods of elastic retraction. However, there was a lack of clear guidance in previous studies regarding the appropriate magnitude of palatal elastic retraction force. We gradually increased the magnitude of palatal elastic retraction force to establish an appropriate magnitude for anterior teeth retraction. As shown in Fig. 4, with a gradient increase in elastic force, incisors in both G2 and G3 exhibited enhanced retraction distance and vertical control, but also an increase in lingual tipping. Due to the critical importance of vertical and torque control in anterior teeth retraction, we recommend that the magnitude of palatal elastic retraction force should be applied between 100 g and 150 g. Magnitudes of force outside of this range should be avoided to prevent excessive pendulum effect of the anterior teeth. Previous finite element studies [2, 9, 17, 27] on anterior retraction often evaluated tooth displacement using either a global or local coordinate system alone. However, this global system failed to provide a detailed analysis of torque control for individual teeth. This strategy may not accurately display the buccolingual tipping or the change of torque when calculating the difference in sagittal displacement between the crown and root [2]. While the local system assessed retraction displacement based on the buccolingual movement of anterior teeth, which is inconsistent with the retraction direction defined in the experimental design [17, 20]. To address this, we analyzed anterior retraction and vertical displacement in the global system and assessed torque changes in the local system. This approach allowed for a more multifaceted analysis of tooth displacement, thereby providing more comprehensive guidance for adding overcorrections in the current CA design system and enhancing clinical applicability. In this study, a significant increase in the maximum von Mises stress of CA in G2 was observed concurrent with the gradient increase of elastic force (Fig. 6). An ideal material for CAs should be sufficiently stiff with high ultimate tensile strength (UTS) to ensure that applied forces remain within the elastic range [28]. In previous studies, the UTS of CAs varies depending on thickness, aligner aging accelerated by intraoral temperature [29], construction material [30], the intraoral environment and manufacturing methods [31]. In their analysis of UTS of thermoformed aligners, Jindal et al. found that the tensile yield stress of 0.75 mm-thick thermoformed aligners was 53.52 MPa; furthermore, aging in human saliva decreased UTS [32]. Furthermore, Kuntz et al. found that intraoral temperature could also accelerate aligner aging and that UTS varied considerably between different brands [29]. In addition, a previous study reported that the mechanical properties of a modified Poly (ethylene terephthalate) glycol/Polycarbonate/Thermoplastic polyurethane blend improved the tear strength to 50.23 MPa [30]. Shirey et al. found that maximum range values for the UTS of thermoformed aligners using EX30 and LD30 (Align Technology) were 45.04 to 71.66 MPa [31]. Therefore, a palatal elastic force of less than 150 g might be considered within the safe range. However, when the force applied at the precision cuts exceeds 200 g, the CA may undergo irreversible deformation or tearing. The anatomical features of the insertion site appear to play an important role in the successful use of mini-implants, including bone quality, characteristics of the soft tissues, cortical bone thickness, and location. The palate naturally offers an excellent location for OMI insertion due to the fact that it is far from the roots of the teeth, and because of the thickness of the soft tissues and the quality of the cortical bone in this area. Therefore, the failure rate for palatal orthodontic anchorage screws is significantly lower than that for buccal screws [12]. Furthermore, some authors took the midline of the hard palate as the origin of coordinates to evaluate the bone mineral density, thickness and soft tissue of the hard palate [33]. The parasagittal area showed an anteroposterior increase in bone density up to the contact area between the first and second molars and then decreasing further posteriorly [34]. In addition, attention should also be paid to avoiding the greater palatine vessels, which are located between the first and second molars on the hard palate, 8.7 to 10.9 mm away from the alveolar crest [18]. However, by measuring the CBCT data of 200 adult subjects, Aldo et al. found that bone volume in the mesiodistal and buccolingual space 5 mm from the alveolar crest, between the palatal roots of the maxillary first and second molars, was sufficient [35]. A site that was 8 mm from the alveolar crest, between the first molars and second molars, was selected as the anchorage screw placement location in this study. However, the impact of different insertion heights of palatal orthodontic anchorage screws on anterior teeth retraction requires further analysis. FEA represents one of the best tools with which to analyze force systems delivered by orthodontic appliances. However, there are several limitations to this study that need to be considered. First, FEA models are only simplified representations of the actual structures [16]. Thus, there was a lack of comprehensive experimental data to characterize the PDL in terms of its non linear, viscoelastic, and anisotropic responses [36]. It should be emphasized that the currently used FEA method could only indicate the initial tooth movement tendencies instead of the long-term remodeling of PDLs and alveolar bone. Secondly, FEA cannot fully replicate the influences of the intraoral environment on the elastic force and aligner force decay, such as the influence of intraoral temperature, saliva, and the frequency of insertion/removal on the CA. During the initial hours of wear, the stress release of the aligner may exceed 50% of its initial stress value; by 24 h after orthodontic load application, the polymer properties of the aligner undergo significant changes, which may affect programmed tooth movement [28]. Fayez et al. found that the longest investigated simulated intraoral application period of 23 h per day induced the largest decay in force, with residual forces after one week of a mere 5% of those initially applied [37]. Finally, this study only analyzed the biomechanical effects of a specific position of the palatal miniscrew on anterior retraction and a single patient model. However, the height of the screw also directly influences the direction of the elastic force and tooth displacement. The model was established based on the CBCT data of a single patient, which lacks population variability. The variability includes crown-root ratio, root morphology, bone density, alveolar bone density, and bone height. Therefore, the results of this study should be regarded as theoretical reference values. Based on the results of this experiment, we will conduct long-term clinical follow-up studies for verification, facilitating translation from research to clinical practice. Conclusions Palatal OMIs combined with lingual traction buttons or precision cuts can effectively increase the incisor retraction distance. Lingual traction buttons are more conducive to the vertical control of the anterior teeth, while precision cuts are beneficial for enhancing anterior torque and posterior anchorage. Palatal OMIs combined with passive metal ligation on the second premolars can also facilitate incisor retraction. However, the tendency for anterior teeth extrusion and lingual inclination increases significantly. The magnitude of the palatal elastic retraction force is recommended to be controlled within the range of 100 to 150 g.