Send Orders for Reprints to reprints@benthamscience.ae 282 The Open Mechanical Engineering Journal, 2015, 9, 282-292 1874-155X/15 2015 Bentham Open Open Access FEM and BEM Stress Analysis of Mandibular Bone Surrounding a Dental Implant Michele Perrella 1 , Pasquale Franciosa 2 and Salvatore Gerbino *,3 1 University of Salerno, Dept. of Industrial Engineering, Fisciano (SA), Italy 2 University of Warwick, WMG, Coventry, UK 3 University of Molise, Engineering Division, Campobasso, Italy Abstract: In the present work the structural behaviour of a mandible with a dental implant, considering a unilateral occlusion, is numerically analysed by means of the Finite Element Method (FEM) and the Boundary Element Method (BEM). The mandible, whose CAD model was obtained by computer tomography scans, is considered as completely edentulous and only modelled in the zone surrounding the implant. The material behaviour of bone is assumed as isotropic linear elastic or, alternatively, as orthotropic linear elastic. With reference to the degree of osteo-integration between the implant and the mandibular bone, a partial osteo-integration is considered; consequently a nonlinear contact analysis is performed, with allowance for friction at the interface between implant and bone. A model of a commercial dental implant is digitised by means of optical 3D scanning process and fully reconstructed in all its geometrical features. Special attention is drawn to the mathematical reconstruction of the CAD model in order to facilitate the meshing process in the BEM environment and reduce the geometrical imperfections generated during the CAD to CAE translation process. The results of FEM and BEM analyses in terms of stress distribution on the mandible are compared in order to benchmark the two methodologies against accuracy and pre-processing efforts. Keywords: BE modelling, dental implant, FE modelling, non-linear contact analysis. 1. INTRODUCTION Endosteal dental implants can cause resorption in the surrounding bone, leading to gradual loosening and ultimately to a complete loss of the implant; in particular a direct correlation was found between overstressed regions and bone resorption [1]. Stress distribution in the bone strongly depends on the implant shape so it becomes of uttermost importance to test different commercial implants in order to devise the configuration providing the lowest possible stress concentration in the bone, thereby reducing the resorption risk. In [2] authors studied the stress distribution on endosseous dental implant and surrounding bone, by using both Finite Element Method (FEM) and Boundary Element Method (BEM) in a three dimensional modelling approach. A FEM-based study in [3] has pointed out the impact of the articular disc stiffness and of the temporo-mandibular joint friction coefficient on the mandible stress peaks and occlusal forces. A later comparison also with BEM was then carried out in [4]. In [5, 6] also the occlusal stress transmitted to the inferior alveolar nerve was analysed. In [7] the biomechanical behaviour, in terms of stress concentration and distribution, of different commercial dental implants with different thread profiles was studied with FEM, *Address correspondence to this author at the University of Molise, Engineering Division, Campobasso, Italy; Tel +39 0874 404593; E-mail: salvatore.gerbino@unimol.it analysing conditions of both perfect and partial osteo- integration. Similarly, by using FEM in [8] a statistical approach was described to evaluate the geometrical parameters of the implant that significantly affect the induced stresses and damage in the bone. All these studies require the modelling of the dental implant and of the surrounding bone topology. In some cases the analysis of simplified geometrical models is sufficient for a preliminary analysis, but for a closer investigation, such that adopted in the present study, detailed geometries of the real models are necessary. The virtual model of the dental implant can be generated by using high density optical scanners providing a dense point data set, which is used to create a tessellated/polygonal model that can be converted into a CAD model by fitting patch surfaces following the common reverse engineering (RE) procedures. On the other hand, the process of virtual modelling of anatomical parts is slightly different: the data set usually comes from 2D medical imaging like CT scan; it is then converted into a tessellated 3D model by volumetric segmentation algorithms. The polygonal model can be edited with RE techniques or, taking into account the specific topology of the part, CAD-like modelling functions (extrusion, sweep, loft, blend) can be used to create the final 3D CAD model [11]. The present work starts from an accurate modelling of a commercial dental implant as well as the human edentulous mandible and is intended for numerically analysing the stress gradient induced by the loaded implant, using both FEM and