Analysing the mechanical behaviour of thermoformed dental clear aligners using acoustic emission technique

Clear aligners have been widely preferred in recent years for aesthetically pleasant orthodontic treatments. However, their efficacy in treating dental malposition and malocclusion is yet to be studied extensively. Clear aligners are mostly made from thermoplastic materials that guarantee sufficient elasticity while lowering the plastic deformations during their use. This is to guarantee a stable level of forces acting on teeth to be repositioned. This work aims to get insight into the mechanical behaviour of these devices during their period of use, by comparing aligners produced from two different thermoplastic materials: polyethylene terephthalate-glycol modified (PET-G) and polyurethane (PU) supplied by Dooris and Ghost Aligners, respectively. The aligners were subjected to cyclic compression loading, to simulate the swallowing act throughout 15 days of use. Tests were conducted by surrounding the aligners with artificial saliva to simulate the intraoral environment. The Acoustic Emission (AE) technique was used to analyse the damage progression in the aligners during the loading. Furthermore, the AE results were compared with the energy absorbed and the stiffness changes in the aligners. Finally, the damage evolution in the aligners post-loading was validated using optical microscopy. The AE results revealed a good correlation with mechanical and optical microscopy data, thus contributing to the understanding of the mechanical behaviour of the clear aligners.


Introduction
In the last 30 years, the number of adults seeking orthodontic treatment has increased significantly.Nonetheless, most adults are reluctant to use conventional treatments with wires, bands or brackets to treat malocclusion [1].Clear plastic aligners have emerged as orthodontic devices to overcome the adult's indecisiveness in undergoing orthodontic treatments.The applications of medical-grade thermoplastic materials in dentistry are well-known [2].Clear dental aligners are made from thermoplastic materials and are used for a short period of time, typically between 7 to 14 days.They are designed to apply incremental forces on the malpositioned tooth or a group of teeth and realign them to the desired position.During the use of these aligners, the teeth are programmed to move in small increments, say 0.1 mm to 0.2 mm, before changing them.Since they are used for a short duration, the risk of periodontal diseases such as gingivitis and tooth decay can be avoided.On maintaining oral hygiene, clear dental aligners are preferred over traditional treatments with wires and brackets [3].1306 (2024) 012018 IOP Publishing doi:10.1088/1757-899X/1306/1/012018 2 Despite their advantages, clear dental aligners are susceptible to deformation due to the swallowing forces, since they are commonly produced from thermoplastic materials.Moreover, prolonged exposure to the intra-oral environment may result in their deformation, but it depends on the hydrophilic tendencies of the type of thermoplastic material used.Apart from these, the manufacturing method influences the mechanical performance, dimensional stability, and force delivery characteristics of the clear aligners [4].Therefore, it is imperative to study the mechanical performance of the clear dental aligners under loading.The study of the mechanical performance of the material for dental aligners is enriched in literature, however, the number of research works on the performance of dental aligners is very few.The goal of this research work is to study the mechanical performance of thermoformed clear dental aligners produced using two different thermoplastic materials PU and PET-G.
In the authors' previous works, the mechanical performance of thermoformed clear aligners was studied and the strain evolution in different dental positions during the cyclic compression load was analysed using the Digital Image Correlation (DIC) technique [5].Although the analysis of mechanical performance was successful, the damage evolution stages during the use of the aligners should not be identified.Therefore, in this research work, a passive Non-destructive Evaluation (NDE) tool, the Acoustic Emission (AE) technique is used to study the mechanical behaviour and damage evolution stages in the clear dental aligners.The AE technique is based on acquiring the stress waves (or acoustic waves) generated from a material under loading.These stress waves are generated when the stored strain energy in the material is released rapidly due to irreversible deformations [6].Analysing the time and frequency characteristics of these stress waves may reveal information about their source.
The objective of this study is to compare the mechanical performances of two clear dental aligners made from PET-G (supplied by Dooris) and PU (supplied by Ghost Aligners), using the AE technique.

Preparation of Clear Dental Aligners
First, the dental alignment of a patient is scanned using a commercially available intra-oral scanner and reconstructed in 3D using commercial software with an accuracy of 6.9 µm.Then, the solid resin casts of the upper and lower jaw are constructed from the scanned output using daylight hard resin.The resin casts are constructed using the Liquid Crystal HR2 3D printer.Medical-grade sheets of PU and PET-G of thickness 0.75 mm are thermoformed over the hard resin casts using Erkoform 3D vacuum machine to prepare the clear dental aligners.The sheets are first heated by a medium-wave infrared heater before applying the vacuum.The respective aligners are named as TPET-G and TPU.

Cyclic Compression Tests
The mechanical tests are designed to simulate the forces applied to the dental aligners during the swallowing act during the period of use of the aligners, which is roughly equal to 15 days or 22500 cycles.The force induced by the upper jaw during swallowing is 50 N [7].Therefore, each cycle of the compression test is designed into 4 stages.In the first stage, the load is ramped from 0 to 50 N in 1 s and in the second, the 50 N load is kept for a dwell period of 1 s.In the third and fourth cycles, the load is ramped down from 50 N to 0 in 1 s and a dwell period of 1 s at 0 N, respectively.
The cyclic tests are carried out on an Instron 3344 single-column universal testing machine (1 kN load capacity).The lower and upper resin casts are firmly held onto the pneumatic vices of the testing machine and the clear aligner is mounted onto the upper cast.In order to simulate an intra-oral environment during the loading, a sponge impregnated with artificial saliva is kept IOP Publishing doi:10.1088/1757-899X/1306/1/0120183 in contact with the aligner throughout the test.The upper and lower casts are held firmly at a contact pressure of 10 MPa before the commencement of the test.

Acoustic Emission Test Setup
The stress waves generated during the compression cyclic loading are recorded using a piezoelectric transducer, PICO sensor (Physical Acoustics).It is a lightweight miniature sensor, which has an operating frequency of up to 750 kHz and has two resonant frequencies: one at 250 kHz and the other at 550 kHz.The stress waves that cross the detection threshold of 26 dB are acquired at a sampling rate of 2 MHz.The signals are preamplified by 60 dB and passed through a flat band-pass filter of 100 kHz to 1 MHz before acquisition.
The frequency characteristics of the stress waves generated from the complex-shaped clear aligners are unknown and not reported in the literature [6].Therefore, a preliminary test is carried out to study their frequency characteristics.A previously tested aligner, which has a small crack in the first third molar is mounted onto the test machine and the compression cyclic load was applied for 100 cycles.However, this time, the loads during the ramping are set as 100 N (instead of 50 N), in order to propagate the crack.The stress waves acquired from this analysis are recorded and their frequency characteristics are reported in the following section.

Preliminary Cyclic Test Results
As reported in Section 2.3, a preliminary test is carried out to understand the frequency characteristics of the stress waves generated from the propagating crack in the dental aligners.The distribution of the acoustic waves in terms of their peak frequency and peak amplitude is presented in Figure 1.A sample AE signal in its time domain and in its frequency domain (calculated using Fast Fourier Transform (FFT)) is presented in Figures 2a and 2b, respectively.Once again, the results show that most of the frequency components lie below the 200 kHz frequency band, while the central frequency is around 150 kHz.

Mechanical Performance of the Aligners
The mechanical performances of the aligners are analysed from the hysteresis loop of the loaddisplacement curve.The total area enclosed by the curve is used to calculate the energy absorbed by the aligner in each cycle and the first slope of the load-displacement curve is used to study the stiffness.The results are presented in Figures 3a and 3b, respectively.It can be observed in Figure 3a that the energy absorbed by TPET-G increases in the initial cycles (up to cycle 4000) and remains stable thereafter.On the other hand, the average energy absorbed by TPU is quite larger than TPET-G up to cycle 17000, and then it drops suddenly and decreases linearly until the end of the test.During the cyclic compression test on thermoplastic materials, the energy absorbed is proportional to the strain deformation in the material during the initial cycles.PU are thermoplastic materials characterized by large plastic deformation, which could be the reason why the average energy absorbed by TPU is greater than TPET-G.The large plastic deformation in thermoplastic material is followed by strain hardening, which results in an increase in stiffness.From Figure 3b, it can be observed that after 10000 cycles, the stiffness of TPU increases exponentially up to cycle 17000, where an energy drop is observed in Figure 3a.
However, the reason behind the sudden drop in the energy absorbed by TPU is quite unclear.In thermoplastic materials, the release of strain energy is associated with the formation and propagation of damage.Since the energy drop did not occur linearly but occurred in a rapid manner between cycles 17000 and 18000, it can be assumed that the damage that occurred was rapid.Based on the stiffness curve in Figure 3b, since the stiffness increases exponentially from cycle 10000, it can be assumed that the strain hardening commences to occur at this cycle and finally results in the propagation of damage at cycle 17000.
PET-G polymeric materials are often characterized by higher modulus and higher stiffness than PU [9,10].This is obvious from both Figures 3a and 3b that the energy absorbed by TPET-G is quite low compared to TPU, while its stiffness is higher.Moreover, both the energy and stiffness of this aligner remain more or less stable throughout the test.This clearly indicates that the performance of TPET-G is better than TPU.

Optical Microscopy Results
In order to validate the assumptions made in the previous section, both optical microscopy and AE results are used.First, the microscopic images of untested TPU aligners are presented in Figure 4. (Unfortunately, the microscopic images of untested TPET-G are unavailable at the moment).The results show that there are some microcracks of length less than 1 mm and some strainhardening regions that can be observed on the external surface of the aligner in Figure 4.These probably are due to the thermoforming process.It is safe to assume that these microcracks serve as the regions of stress concentration, which results in the damage progression in these aligners.To verify this, the microscopic images of the tested TPU and TPET-G are presented in Figures 5 and 6, respectively.4 of the untested microscopic images, there are regions of strain hardening and microcracks of length less than 1 mm that can be observed in Figures 5a, 5b and 5c.However, in Figure 5d, several larger cracks with lengths greater than 1.15 mm and up to 3.69 mm can be observed.Moreover, several whitening regions of beachmarks indicating fatigue failure due to the cyclic load are also observed.This validates the assumption that the energy absorbed by TPU is released due to the formation of these cracks.Based on the stiffness curves and energy curves in Figure 3, it can be confirmed that this occurred at cycle 17000.
The microscopic images of the TPET-G aligner post-testing are presented in Figure 6.As expected from the mechanical results, only one major crack is observed in Figure 6a, whose length is significantly smaller than the one observed in TPU.Besides, there are no beachmarks around the failure, indicating that these are not due to the fatigue cycle.Most of the other microcracks and micro-failures in Figures 6b, 6c and 6d indicate the substantially good performance of TPET-G compared to TPU.
The microscopic images shed light on the regions of failure, the type of failure and the occurrence of failure when compared with the mechanical performance of the aligners.Despite identifying the occurrence of the final failure, an in-situ analysis is essential to understand the commencement of these failure modes.For this, the AE results are presented and discussed in the next section.

Acoustic Emission Results
The AE results are compared with the mechanical test results in this section.The time domain parameter of the AE signals, cumulative counts, which indicate the number of instances the amplitude of the stress waves crosses the detection threshold is compared with the energy absorbed by the aligner.The results for TPU and TPET-G are presented respectively in Figures 7a and 7b.
The first obvious difference between TPU and TPET-G can be found in the maximum number of cumulative counts recorded in each aligner.In TPU it is about 10500, while on TPET-G it is only 4750.This clearly shows that the number of stress waves generated in TPU is more than that of TPET-G.This validates the assumptions made in both the mechanical test results in Section 3.2 and the optical microscopy results in Section 3.2.
Secondly, it gives an indication of the onset of the damage in each of these aligners.For example, in Figure 7b, it can be seen that the cumulative counts remain close to 2 until cycle 7000 and it increases rapidly beyond that point and then it stabilises.This indicates that the crack observed in Figure 6a starts to propagate at this point and then it stabilises.Since there is no major change in the energy absorbed by this aligner during this phase, it can be concluded that this damage did not affect the mechanical performance of the TPET-G aligner.
Thirdly, from Figure 7a, the onset of different damages can be examined in the TPU aligner.The first damage onsets as early as cycle 2000 and it progresses until cycle 5000.Then there is a period of stabilisation up to cycle 9500.Beyond that cycle, the cumulative counts begin to increase gradually and it extends until the end of the test.The mechanical results indicate that the energy starts to drop from cycle 11000 but the AE results can precisely indicate the onset Finally, in order to verify that these AE signals are crack growth events, two signals from each aligner are randomly chosen and their frequency characteristics are studied using FFT.The results are presented in Figure 8.The AE results clearly show that the major frequency components of the AE signals are centred between 100 kHz and 200 kHz, which was observed during the preliminary tests in Section 3.1.This confirms that these signals are generated from the crack growth events in the aligners.

Conclusion
The mechanical performances of two different thermoplastic clear dental aligners are evaluated and compared in this research work.The mechanical results show that the aligner thermoformed from PET-G has stable energy absorption and stiffness characteristics compared to that of PU.The AE technique is used to identify the damage evolution in these aligners under cyclic loading.The results are promising, particularly, the time-domain-based and frequency-based analyses.The time-domain-based shows the onset of the damage progression in the aligners, while the frequency-based analysis shows the type of damage.The results show that this technique has a greater potential in the evaluation of the mechanical performances of orthodontic devices.

Figure 1 .
Figure 1.Distribution of Acoustic Waves as a function of their Peak Frequency and Peak Amplitude

Figure 2 .
Figure 2. (a) Sample AE Signal in Time Domain and (b) Frequency Domain

Figure 3 .
Figure 3. (a) Energy absorbed and (b) Stiffness variations in the aligners during the Compression Cycles

Figure 5 .
Figure 5. Microscopic Images of Tested TPU Aligner taken at different Tooth Positions

7 Figure 6 .
Figure 6.Microscopic Images of Tested TPET-G Aligner taken at different Tooth Positions

Figure 7 .
Figure 7. AE Results compared with Mechanical Results for (a) TPU and (b) TPET-G Clear Dental Aligners

Figure 8 .
Figure 8. Frequency Characteristics of Randomly Selected AE Signals (a) and (b) from TPU; and (c) and (d) from TPET-G