Browsing by Author "Martin, John"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Open Access
    Modelling directional casting processes in which heat conduction and cavity radiation are the dominant modes of heat transfer
    (2002) Kotschy, P J; Mitchell, G; Nurick, Gerald; Martin, John
    Directional investment casting processes involve complex interactions of various mechanisms of heat and mass thansfer in spatially complex domains and in the presence of a change of phase. In particular, the transfer of heat within the furnace occurs in the form of conduction, convection and radiation. This thesis addresses the development of computational techniques to simulate, at a macroscopic scale, such casting processes. In this study the conservation of heat energy within the casting is assumed to be maintained by conduction, accompanied by the release of latent heat energy during solidification. The overall state of the radiation in the furnace chamber is analysed in terms of the absorbed, emitted and reflected energies for each surface defining the geometry of the chamber. By applying a diffuse-grey approximation to these energies, an expression for the net flux for each surface is derived.
  • Thumbnail Image
    Item
    Open Access
    Numerical simulation of implant-bone interaction following cementless joint replacement
    (1996) Starke, Gregory Richard; Martin, John; Spirakis, Thanos
    The advent of cemented joint replacements has revolutionised the management of patients suffering from chronic arthritis. However, establishing a durable bond between the prosthesis and the surrounding bone remains a considerable problem. As a result, cementless implants have been developed. These components rely on the ingrowth of bone into a porous coating, which covers a portion of the component surface, to achieve the required mechanical interlock. Once mineralised bone tissue has formed within .the porous surface, a stable bond results which will be maintained by the normal bone turnover processes, thereby providing long term attachment. However, one of the problems associated with the use of cementless implants is the unpredictability of the extent of bone ingrowth. The process of osseo-integration is greatly influenced by the magnitude of the micro-motion between the implant and the surrounding bone. Large movements inhibit ingrowth, and may result in the formation of an interfacial fibrous tissue layer. In addition, interface strains will influence the early repair process and guide long term bone remodelling within this region. A numerical model for the prediction of bone formation within the porous surface has been developed. The evolution laws consider the early repair activity, possible fibrous tissue formation, and long term remodelling, as a function of the history of inelastic relative displacements and elastic interface strains. The model is based on the development of an isoparametric interface element, which is suitable for implementation into a non-linear finite element code. In the unbonded condition, the contact is governed by a Coulomb friction formulation. The position and shape of the Coulomb yield surface is altered according to the evolution equations, which govern the development of mineralised tissue within the surface porosity. The strain history and post-operative time are then used to develop a stimulus coefficient, which determines the course of the interface tissue development. If bone tissue is predicted, the subsequent interfacial material will be governed by a remodelling algorithm for the prediction of the long term response. If the bond strength is exceeded, fracture occurs and the joint may open or slide, thus returning to its original, unbonded, state. In the event of large micro-motions, the yield surface and material formulation are altered to include fibrous tissue. The model is used to predict the development of interfacial tissues at the porous surface of a tibial tray component, with a central peg and a PCA (Howmedica, Inc.) femoral stem. Although many factors influence interfacial tissue development, mechanical loads are assumed to be dominant. In the short term, the relationship between micro-motion and interface tissue response has been shown. However, long term remodelling of interfacial tissues has not been widely demonstrated and, therefore, additional experimental data is required to validate the current long term remodelling predictions.