Implant retained overdenture offers a convenient treatment for edentulous patient. There are different types of attachment used to retain implant-overdenture such as studs, bars or combination form. Recently, the profession had a chance to study both the acceptance and the success of implant-retained overdentures versus fixed appliances. Overdentures are not only an acceptable restoration, but also the restoration of choice in many instances.1,2

For several years, both bar and solitary stud systems were used as attachment systems with different designs and applications. Some clinicians preferred using stud systems due to their simplicity and suitability in many clinical conditions.3,4 The use of extracoronal resilient attachment (ERA), as implant overdenture attachment, is well documented in conventional dentistry. ERA has also broad applications for removable partial dentures and tooth-supported overdentures. Their specific restorative design, by exchanging position of male and female potions, acts to affect the load distribution to the supporting implants. Moreover, it enables less supporting implants needed, which is economic for many patients.3-7

Many researchers used to treat their patients by implant overdentures retained and/or supported by many attachment systems. These attachments act through splinted or solitary designs. Solitary systems, such as ERA and Zest Anchors, act at minimal interarch restorative space, beneath incisal or occlusal surface of the prosthesis, than any splinted anchorage designs. Although non-paralleled implants may deteriorate retention of solitary stud systems, solitary systems are hygienic, cheap and their technique is more convenient.8 Furthermore, if bar systems were used, they act to inhibit displacing forces in both vertical and oblique directions. They also tend to behave as one unit and so they are more retentive and stable than solitary abutments.9-11

There is no accessible clinical method to study stress-strain distribution of implant overdentures at implant-bone interface. However, researchers were able to study stress analysis using strain gauge only at the abutment level.12,13 On the other hand, photoelasticity and the finite element analysis (FEA) are powerful simulation methods used in dentistry. They have enabled better understanding of stress transmission and distribution at implant-bone interface.14-16

Authors used FEA broadly to study the influence of different attachment on the bone integrity of implant-overdenture.17-22 Vafaei et al.23 examined the effect of overdenture attachment design on the strain distributions and values of both bone and implants. Results from the bar design showed smaller strain magnitudes in both laterotrusive and protrusive motions. Thus, they claimed that, bar design was considered superior than the single standing implants with attachment. Furthermore, Tabata et al.9 compared the stress-strain analysis of splinted versus non-splinted implant overdenture. They found that the use of single standing implants with attachment induced more stress in bony tissues than the bar-clip system.

Federick and Caputo24 conducted a photoelastic study to compare different attachment designs including splinted and non-splinted ERA system. They concluded that ERA implant abutment provided the most equitable force transfer, as evidenced by more uniform stress transmission from implants to bones. Additionally, Daas et al.25 performed 3D finite element analysis in implant-overdenture based on non-linear material setup. The results clarified the favorable role of attachment resiliency in reducing stress concentration around implants.

Moreover, Liu et al.26 and Osman et al.27 conducted studies on implant-retained overdentures to reveal stress-strain analysis in their design. They assumed movement and sliding between overdenture and attachment units during load applications considering frictional contact to simulate such motions.

In an attempt to get the benefits of the Hader bar system and keep the advantages of the ERA system, a combined bar-ERA was designed and mechanically tested. Although, this design could influence the stress-strain condition of the implant-overdenture especially at the bone-implant interface area. Therefore, the aim of the present study was to assess the influence of using different attachment system on stress-strain distribution of two splinted-implant overdentures. The first overdenture was retained by Hader bar clip whereas the second overdenture was retained by two ERA studs placed distal to Hader bar bilaterally. The maximum equivalent stress and strain values and their distribution were studied at the bone-implant interface level using non-linear contact 3D finite element analysis method.

3D model of an edentulous mandible was created using SolidWorks software (SolidWorks Corporation, Concord, MA, USA) and nurbs modeling software MOI software (MOI v 2, Triple squid software design, USA). The actual dimensions, cross section and configuration of the 3D model were copied from cross-sectional and horizontal cut images of a previously CT-scanned edentulous patient (GE Medical System/Bright Speed S, USA). Modeling process was started by selecting six properly distributed cross-sectional images bilaterally (two at molar area, two at canine area and two at incisor area). These images were used to create sketches where bone border was traced. After tracing, loft operation was performed to create the bone volume guided by the arch outline as seen from the horizontal view. After creating 3D model of the cancellous bone, the compact bone was produced as a shell of variable thicknesses between 1-2 mm around cancellous bone. Mucosa was also modeled as a layer of 2 mm thickness covering compact bone. In addition, two simplified implant fixtures with their attached bar abutments were designed to simulate dimensions (3.3 mm × 13 mm) and shape of the actual implant (Brånemark System Mk III; Nobel Biocare AB, Göteborg, Sweden). Two implants were virtually placed in canine area bilaterally using (SolidWorks). There are two attachment systems designed for twostudy models:

The finite element method was selected as a stress-strain analysis method that enables accurate, feasible and satisfactory results for many dental applications. Finite element is a numerical stress-strain analysis method used to analyze the assemblies of the studied designs. Therefore, Ansys software (ANSYS Workbench v 14 package; ANSYS, Inc., Canonsburg, PA, USA) was the software of choice to perform the study using its graphic user interface (Ansys workbench) and static structural was the type of the analysis selected. The software is classified into five modules (design modeler, engineering data, model setup, model solution and results). Thus, the two assemblies were transferred from SolidWorks to Ansys Modeler module of the finite element software. All parts were divided into small components representing element with a process called meshing. Meshing process was performed using 3D tetrahedron element type with four-node element shape, Fig. 2. The total numbers of elements and nodes are listed in Table 1.

The different material properties were assigned to each part according to Table 1. All material properties were assumed linear, homogenous and isotropic to facilitate calculation process and reduce solving time. Accordingly, two values (Young's modulus and Poisson's ratio) were assigned for each part, Table 2.10,14,18,20,27

The contact behavior among different parts of the models was set so that it is bonded by surface-to-surface contact. In order to allow sliding and movement between overdenture and mucosal surface, a non-linear frictional contact type was assumed. The coefficient of sliding friction between the overdenture surface and mucosal surface was set to 0.334.26 In addition, a no-penetration sliding (friction coefficient = 0.3) contact was defined for attachment components.27

The models were constrained at their rami ends bilaterally with zero degree of freedom at these areas, rather than using inferior border, to permit mandibular flection and so simulating the real behavior, Fig. 3. A 200 N unilateral static load was applied perpendicular to the occlusal plane (occlusal surface of the overdenture) in the molar area of the left side of both studied designs, Fig. 3.17

The stress-strain distribution of the von Mises stress and strain were computed and analyzed at the bone-implant interface from both sides (implants and bone) of both models. The values of the maximum equivalent stress and strain were then recorded and charted and interpreted.

The von Mises stress and strain values and their distribution were generated by the finite element software according to a stress and strain map with a color scale (lowest stress values = dark blue; highest stress values = red). In addition, quantitative analysis was performed using measurements of both the stress values (in megapascals) and the strain values in regions of bone-implant interface of each group. These values were converted into graphs to facilitate data interpretation.

The values of maximum equivalent stress of bone at the implant-bone interface around both fixtures in loaded side (F1) and unloaded side (F2) were recorded. The stress-strain distribution was studied on the bony surfaces at the bone-implant interface of both implants (F1 and F2). The maximum value recorded for bone around F1 at bone-implant interface was (6.4221 MPa). This value of maximum equivalent stress was recognized at the crestal bone tissues especially from lingual side, Fig. 4A. In addition, various areas of stress concentrations were also seen at the buccal plate coronally and lingual plate of bone apically.

The bone around implant F2 showed a maximum value of equivalent stress equal (5.2119 MPa). This area of stress concentrations was seen also at crestal bone from the lingual side, Fig. 4B. Certain areas of stress could also be noticed at outer plate of bone bucally and lingually.

Regarding implants surfaces, the implant of the loaded side exhibited a maximum value of equivalent stress (2.4974 MPa). On the other hand, a higher value of stress concentrations (5.3193 MPa) was recorded on the surface of the second implant at lingual side of the first thread, Fig. 4C.

Upon analyzing stress concentration in bone around implants of the second model, an area of stress concentration (8.1509 MPa) was noticed at the level of the first thread of the F1 implant both buccally and lingually. Other areas of variable values of stress were also seen at the crestal bone and cortical bone shell of the apical lingual plate, Fig. 4D.

The bone around F2 showed a highest value of stress (8.1757 MPa) which could be recognized at the level of the first thread either buccally and lingually. Another area of stress could also be seen surrounding implant neck at the crestal bone and lingual cutback of implant body, Fig. 4E.

The maximum equivalent stress on the implants surface was (18.624 MPa) for the implant at the loaded side F1. The other implant F2 showed stress equal (31.77 MPa), as seen in the cutback area of F2 body, Fig. 4F.

The maximum value of the stress was recorded surrounding F2 bone of the second model, followed by the contralateral side bone, then bone surrounding F1 of the first model and finally the bone surrounding implant F2. Implants also showed a highest value of stress for F2 of the second model, followed by F1 of the same model, then F2 of the first model and the least value for F1 of the first model, Fig. 5.

Maximum equivalent strains in different models structures were collected and studied. The strain distribution could also be followed using the color scale that represents different levels of strain values.

The bone around implant F1 revealed a strain value equal (5.7761 × 10^-4) which could be noticed at the lingual side of the crestal bone. Generally, there were several strained areas of lower values could also be seen around many threads and the cortical plate of bone lingually, Fig. 6A. The maximum equivalent strain recorded for bone around implant F2 was (4.9748 × 10^-4). This value represented by strain collected at the lingual side of the crestal bone. There were also several strain areas both buccally and lingually and also around implant threads, Fig. 6B.

The implant F2 showed a strain value equal (5.5527 × 10^-5) which could be recognized clearly at the implant first pitch. However, areas of lower strain values were also seen around several threads and implant's neck. On the other hand, the implant F1 exhibited strain value (2.6225 × 10^-5) as seen at the implant first thread, Fig. 6C.

Generally, maximum equivalent stress values were recognized in the bone around F1 and F2 at the level of the first implant thread and at apical third near the implant cutback with values equal to (5.8824 × 10^-4) and (6.507 × 10^-4) respectively, Fig. 6D and Fig. 6E.

Implants strains values were (2.2015 × 10^-4) and (3.5987 × 10^-3) for implants F1 and F2 respectively. However, the maximum value recorded for F1 was recognized near the implant collar, the maximum value for implant F2 was seen at the implant cutback area apically, Fig. 6F.

The highest values for maximum equivalent strain in Model 2 were recorded at implant F2 followed by bone surrounding it, then bone around F1 and finally implant F1. Regarding Model 1, the highest values for maximum equivalent strain in bone-implant interface was recorded around F1 bone, followed by bone around F2, then implant F2 and finally implants F1 as seen in Fig. 7.

Once overdenture was suggested as a treatment for implant patients, researchers studied many attachments to select the most suitable system.10-15 Accordingly, Misch28 suggested many attachment designs when selecting implant overdenture as a line of treatment. The remaining bone volume, number and position of implants, and expected prosthesis movements influence the criteria of the selected attachment. Consequently, they added that number and distributions of these attachments might also affect the integrity of the design.

Regarding the present study, although hader bar-clip system prevents independent implant movements, splinting may produce a favorable effect on stress-strain distribution. When the hader bar-clip system was selected for implant overdenture patient, the expected movement from this design was true hinge movement around single axis of rotation.2,29 This hinge limits movement into two directions. Whereas this overdenture is implant-retained, it is considered tissue-supported. As a result, posterior residual ridge bears the majority of force and sacrifice bone anteriorly.28-30 In addition, mucosa resiliency has an influence on prosthesis movement amplitude. This might coincides with the present study to explain bar-clip model findings. Bar-clip model generally exhibited favorable stress-strain distribution at the bone-implant interface. It was clearly recognized that both values of maximum equivalent stress and strain were lower than their counterpart of the other model. This was in agreement with Tabata et al.,9 as they showed a better stress distribution of the bar design over both mucosa and cortical bone. In addition, Bergendal and Engquist31 showed, in a prospective clinical study, that the implant loss using bar-retained overdentures was less than any other attachment type used in their study.

Generally, in model one, the crestal bone and the coronal part of the implant were the areas of stress concentration and high strain response. These findings matched the overall investigations regarding crestal bone both clinically and using FEA.32,33 These authors focused their investigation on the crestal bone as area of frequent bone resorption and apparent stress concentration as seen for bar-retained implant overdenture.

On the other hand, the ERA-bar attachment showed higher values for both stress and strain than model 1 at the bone-implant interface. Unfortunately, adding ERA distal to the bar showed a stressful condition to the implant-overdenture and stimulated stress at the unloaded side. These finding was in agreement with Ochiai et al.15 study during simulating the laterotrusive movement. On the other hand, Celik and Uludag5 reported lower stress values for bar era combination compared with conventional bar-clip attachment system. The ERA-bar attachment looks to behave another approach upon applying the force. By adding two era attachment as a mean of retention distal to splinted implants, the system became more complicated. The use of distal attachments on bar-clip system creates a fulcrum line at this portion. The prosthesis rotates in the sagittal plane around the fulcrum axis and due to the elastic modulus of the resilient matrices which fits the attachments; the stress magnitude on the implants was reduced.22

The actual overdenture movement may be completely different from their attachments when they are independent. In addition, prosthesis movement could be changed from one to six directions using the same type of attachments.28,34 Furthermore, when the applied force was only assigned to bolus side, as assumed in the present study, the position of the fulcrum line might be changed. The fulcrum line is the line at which the displacement was zero. Moreover, in class I lever the resistance, found on the other side of the fulcrum, is opposite to the force direction.28,34-37 Accordingly, when the force was applied unilaterally in the ERA model, the ERA attachment might act to suppress overdenture movements distal to the first implant, despite of their resiliency.28,36 Thus, the ERA attachment near to the applied force tends to stop the movement and the movement there will be zero. At this level, the ERA near to the applied force became the fulcrum and hence the force was interacted with the resistance at the opposite side of the fulcrum. As a result, the resistance was assumed to displace the ERA matrix from the resistance side and so the apical part of the bone around the second implant may interact in an opposite direction and revealed a high stress value (8.1757 MPa). This hypothesis could clarify the stress concentrations seen in the apical third of the second implant near the cutback area. Alternatively, stress concentration was seen near collar area of the bar-clip model.

The contact management could influence the results of this study. The contact between implants and bone linear bonded contact was assumed representing 100% osseointegrations as expecting from delayed loading protocol.9,10,14,16 Although, sliding and movements between mucosa and denture and also between different attachment components were set to non-linear frictional contact to simulate different movements expected.26,27

There were certain simplifications rather than limitations were assumed. Regarding modeling of the mandibular bone, both coronoid and condyloid processes were neglected from design. The restrains were assumed at the rami ends and not from proposed muscle attachments.25 The applied force was static and not dynamic to simulate chewing cycle.38

Within the limitation of this study the following conclusions could be drawn: