Cortical Bone Fracture and Orthopaedic Fixation Devices: An Experimental and Computational Investigation

Abstract

An experimental and computational investigation of cortical bone failure mechanisms has been conducted in this thesis. Firstly, a computational comparison of four methods of proximal humeral fracture fixation was conducted. Peak stresses were predicted at the screw cortical bone interface. Carbon fibre reinforced PEEK (CFPEEK) devices were proposed as an alternative to existing metallic devices. It was demonstrated that CFPEEK devices lowered stresses at the screw cortical bone interface, thus lowering the risk of screw pullout/pushout. Next, a novel experimental test rig was developed, so that crack patterns during screw pullout could be identified in real time as the screw was removed from the cortical bone. Pullout tests were conducted with; (i) osteons aligned parallel to the central axis of the screw (longitudinal pullout) and (ii) osteons aligned perpendicular to the central axis of the screw (transverse pullout). This experimental study uncovered for the first time, the relationship between the microstructural alignment of cortical bone, the pullout strength and the crack patterns. Two methods of computational modelling were subsequently developed to capture the relationship uncovered during the experimental screw pullout study. The first method of element deletion required the use of a phenomenological biphasic multi-layered composite model. This model accurately predicted both the pullout force and crack patterns for longitudinal and transverse pullout. The element deletion method was limited to 2D simulations therefore an alternate method with a lower computational expense was investigated. The second method involved the development of anisotropic damage initiation criteria in conjunction with the extended finite element method (XFEM). In this case, it was not necessary to explicitly represent the geometric microstructure of bone thus lowering the computational demand. This model accurately predicted the relationship between the osteon alignment, failure forces and crack propagation orientation for mode I, mode II and mixed mode loading. Application, of this fully calibrated anisotropic damage XFEM predictive framework, to screw pullout and 3D simulations of proximal humeral fracture repair, highlighted the potential future application of this method in the field of orthopaedic device design.