Stress and strain distributions on short implants with two different prosthetic connections – an in vitro and in silico analysis

Objetivo: Uma biomecânica ideal que minimiza a tensão entre implante e osso pode proporcionar sucesso para implantes osseointegrados. Este estudo avaliou a concentração de deformação no tecido circundante e a tensão nos componentes de dois implantes com diferentes conexões protéticas através de métodos in vitro e in silico. Material e Métodos: Vinte blocos de poliuretano foram divididos em dois grupos (n = 10), seguido da instalação de hexágono interno (IH) (AS Tecnologia Titanium Fix, São José dos Campos, Brasil) ou de implantes cone morse (LT) (Bicon Dental Implants). Para o método da extensometria (SG), foram colocados quatro sensores ao redor dos implantes. Para a análise por elementos finitos (FEA), o mesmo bloco foi modelado e analisado. Foi aplicada uma carga axial (30 kgf) para ambas as metodologias. Os valores de tensão e deformação foram analisados quanto à correlação com o SG. Resultados: Para SG, LT apresentou uma média de deformação mais agressiva (-932) que IH (-632). Para FEA, a LT mostrou menor tensão (-547) que IH (-1169). Conclusão: Para os dois sistemas implantes, os valores de microdeformação capazes de induzir remodelação óssea indesejada não foram medidos. No entanto, para o implante IH, a presença de um parafuso de retenção tem a desvantagem de concentrar a tensão, enquanto um pilar sólido dissipa a carga axial através do implante, o que sugere um melhor desempenho para o grupo LT. ABsTRACT


INTRoDuCTIoN
T he advancement of implantology is owed to the success of osseointegration processes, however, it still presents assembling challenges [1][2][3].The knowledge of the masticatory mechanism forces on the system of prosthesis over implant is crucial for failure prevention [4][5][6].When an implant-supported prosthesis is submitted to a certain load [6][7][8], it promotes bone remodeling [6,9].
From a biomechanical point of view, the connection between abutment and implant must minimize the stress generated at the implant/bone interface [10], avoiding fatiguepromoted micro strains and consequently, bone resorption [11].Another factor that promotes overloading is the use of short implants, as the crown/implant correlation is unfavorable [12] and provides a larger vertical lever arm.
The Finite Element Analysis (FEA) is a good alternative to help on the understanding of stress generated in the masticatory system since 1970.Firstly, using a 2D model [6,13] and from the 80's until today, 3D models [2,3,14] have been used to develop and improve this tool for the study of biomechanical behavior of prostheses and implants.
FEA is the most affordable tool which simulates the same possible damages and presents similar results of an in vitro study [15,16].In FEA, a mathematical model is used to envisage ideal conditions.However, in some situations, the correlation with an in vitro study may be necessary for further results [16,17].The electronic strain [3,18] and the photoelastic methodologies [19] may add as complementary results to FEA.For numerical stress values, the electronic strain gauge seems a good option [20].The purpose of this study was to understand the influence of two different prosthetic connections on stress distribution and micro strains surrounding short implants.
For LT group, the blocks received locking taper implants.The perforations were made with 400:1 and 50 rpm (Bicon Dental Implants).An implant insertion device was used to place the 4.5 x 8 mm implants in the polyurethane blocks.The 10 mm height abutments were placed on the fixture with the system's dental mallet.
For IH group, the blocks received IH implants of 4.5 x 8.5 mm (AS Technology Titanium Fix, São José dos Campos, Brazil).The perforations were made using a progressive sequence of drills at 1800 rpm, and the insertions of implants were performed with 14 rpm with a torque of 40 N/cm.On the fixture's seating platform, the 10 mm height abutments were screwed with a torque of 20 N/cm, using a progressive mechanical torque wrench.

Strain Gauge Analysis
The strain gauges (SGs) model PA-060-120-L-040AB (Excel Sensors Ind. and Export Ltd., Embu, Brazil) were bonded on the surface of the polyurethane blocks with a thin layer of cyanoacrylate based adhesive (Loctite Super Bonder; São Paulo, Brazil).Four SGs were placed diametrically opposed and tangential at 1 mm around each implant (Figure 1).Terminal plates, responsible for the electrical connections, were attached on the blocks' external surface, connecting the SGs to an electrical signalconditioning unit (Model 5100 Scanner unit-System 5000B; Instruments Division Measurements Group, Inc., Raleigh, NC).Then, a static vertical load of 30 kgf was applied during 10 s using a customized load application device [3].The device's spherical tip was positioned on the center of each abutment.The magnitude of strain on each SG was recorded in units of micro strain (µm/ µm).

Finite Element Analysis
During pre-processing with Rhinoceros software (version 4.0 SR8McNell, Seattle, USA), a solid block was built to simulate the polyurethane block.The models were created with the same size of the blocks used in strain gauge analysis.Then, the evaluated areas were simulated on blocks' surfaces.LT implants were individually modeled following manufacturer's measurements (8.5 x 4.5 mm).Solid abutment were designed with 10 mm without shoulder and shows the slopes coming from the connection locking taper with walls diverging in the angle of 1.5°.The Boolean difference, a Rhinoceros command that trims the shared areas of selected polysurfaces with another set of polysurfaces, was made to the connection's internal walls that was in intimate contact with the abutment's external walls (Figure 2a-2g).
For the IH group, the implant was modeled with 8.5 mm height and 4.1 mm in diameter, with similar internal and external geometry threads.The abutment featured 10 mm height and 4.1 mm in external diameter base tapering to 3 mm at the upper end.A real size screw, responsible for the union with the implant, was modeled with an internal hexagon of 1.2 mm in diameter to simulate the key input torque and similar geometry threads along its length (Figure 2a-2g).
The final geometry was exported in STEP format to ANSYS software analysis (ANSYS 15.0, ANSYS Inc., Houston, USA) (Figure 2h-2k).All materials were considered homogeneous, isotropic, linear and elastic.Respective elastic modulus and Poisson's ratio were designated (Table 1).The mesh convergence test was used and the ideal size of the elements was 0.3 mm.The number of nodes and tetrahedral solid elements were respectively 210,334 and 122,275 for Locking Taper and, 268,363 and 154,738 for Internal Hexagon.All contacts were considered completely bonded, excluding any loss of torque or rotational misfit.

ResulTs
The experimental setup followed in vitro test, 30 kgf axial load on each abutment (Figure 2l-m).The gradient stress generated in the polyurethane block was quantitatively analyzed for total strain (Figure 3), normal elastic strain (Figure 4), maximum elastic strain (Figure 5) and maximum principal stress 6).

DIsCussIoN
Materials that are capable to simulate bone tissue are beneficial due to the reproduction of mechanical properties that allows results similar to the natural tissue [23].Studies using bone tissue may present altered results according to the type of specimen and bone condition.Furthermore, when the focus is the ability to transmit masticatory forces by a system of complex geometries, it is necessary to use simplifications to achieve consistent results before biological studies [24].The use of a healthy bone tissue could never mimic all the factors of human physiology.Thus, a material that simulates bone tissue could respond most questions arising from laboratory routine [25].Polyurethane is an already validated material [23] on international standards ISO 14801: 2007, widely used for mechanical studies with implants [3,26], and it stands out for its human bone simulation capabilities, such as its elastic modulus is between the cortical and alveolar bone [23].
In FEA methodology, an ideal situation is simulated which may be a parameter with initial qualitative results to check what would be the ideal measurement of any instrument in the assembled system.In the logic of mathematical analysis, during a total strain, dissipated energy flows through geometries consistently.In Figure 3, the behavior between groups showing the coherence of the system.In all 20 polyurethane blocks, the biomechanical behavior follows a pattern with statistical difference between gauges in two positions: SG1 and SG3 in X axis, SG2 and SG4 in Y axis.This can be expressed in Figure 6, where the stress distribution in positions 1 and 3 showed higher values while the strain in regions 2 and 4 presented lower values.However, even with a similar strain distribution pattern, values showed a standard deviation that reinforces the need to perform a statistical analysis to better observe the correlation between the behaviors of gauges.
The gauges admeasurement is millimetrically sensitive to any variations, such as installation, minimal surface irregularities, gavel abutment installation friction, chemical interaction between the bonding agent and the transducer surface.These are factors that influence measuring, a situation which often arises from the applied methodology [3,10,17].
To validate this study, the values of strain obtained by in vitro analysis were compared with the results calculated in FEA [16].For this, the gauges area was virtually represented by a square, modeled on the block surface.
In Figure 6, it is possible to observe that the strain distribution is very similar for both methods which validate this 3D analysis.When two different methods are used to analyze the same situation it is possible to separate what should be considered as expected behavior for clinical extrapolation and to avoid erroneous results conclusions, although, it is not significant to be considered as consistent [3,16,18].
With the certain of a valid model, a comparison between implants types could be made.Through a sagittal cut, it may be observed that the maximum principal stress (Fig. 4) is shown in warmer colors which allow verification of the close relationship between implant and polyurethane.Thus, corroborating with other studies that used photoelasticity as method [10,19].Such studies showed lower fringes of strain in conical implants when compared with hexagon connection system.Both models were very similar in values and isolines observed around the fixation but the LT showed mostly stress concentration in a cervical area and IH showed mostly stress in the apical direction.
To clarify the meaning of the values found and explain the difference between groups, it is necessary to verify the stress on the rigid structures of the specimens (Fig. 5): the abutment and implant set.Analyzing the maximum stress generated in implant's connection, the retaining screw in the IH (Fig. 5) showed higher stress concentration.Since this screw is responsible for the connection between abutment and implant, this can make a IH system more fragile than LT.The impact on the threads of this screw tend to be high and it is one possibly reason to justify reports about loss of torque or even fracture of this piece [3,27].LT implant platform presents higher values of micro strain in bone surface (Figure 4) but shows stress in the abutment\ implant joint when compared with the IH connection implant (Figure 5).
Another important factor is the micro strain generated values in both methodologies (Figure 6).In situations in which the connection system inferred a difference between the average strain gauge's values, it is clear that the pathological limit of an unwanted bone resorption has not been reached [9].Nonetheless, bone complication that may occur in these implants and do not appear to be associated with the type of prosthetic connection, but the wrong application/use and surgical planning.

Figure 1 -
Figure 1 -3D model and the experimental model for the LT group.a) 3D model with measurements area, b) Experimental model with strain gauges bonded in the same areas.

Figure 2 -
Figure 2a-m: a-g: 3D models of two-implant system.a) Solid abutment of LT system, b) LT implant, c) LT implant connected with abutment, d) Retention screw of IH system, e) straight abutment, f) IH implant, g) IH implant connected with abutment and retention screw; h-k: Mesh generated for the two implant systems.h) Implant LT placed into representative polyurethane block, i) Implant IH placed into representative polyurethane block, j) Mesh evidenced in abutment and implant LT, k) Mesh evidenced in abutment and IH implant; l and m: Load condition.l) IH implant with load condition in center of abutment and m) LT implant with load condition in the solid center of abutment.

Figure 3 -
Figure 3a) Total strain showing the coherent energy dissipation in LT group, b) Total strain showing the coherent energy dissipation in IH group.

Figure 5 -
Figure 5a) Maximum principal stress (MPS) in IH implant system, b) MPS in LT implant system, c) MPS inside the IH implant system, d) MPS inside the LT implant system, e) MPS inside the IH implant system without abutment and retention screw, f) MPS inside the LT implant system without the solid abutment.

Figure 6 -
Figure 6 -Graph of micro strains for LT and IH groups.

Table 1 -
Distribution of materials mechanical properties.