Numerical modelling of the mechanical behaviour of a fully bio-resorbable glass fibre reinforced composite

  • Xi GAO

Student thesis: PhD Thesis


Since the advent of sterile surgical procedures, load-bearing metallic implants have been used in surgical procedures for bone fixation purposes. The traditional metallic bone fixation devices, i.e. screws, plates or rods, allow doctors to internally set and stabilize fractured bones, meanwhile, however, are commonly associated with the `stress shielding' effect, which weakens the surrounding bone and increases the risk of re-fracture after the device has been removed. The `stress shielding' effect is caused by the fact that the metallic implants have stiff and stable properties in vivo which is beneficial at the early stage of bone healing because the external load is sustained by the implant and fracture bones are stabilized, however, the fracture bone needs to bear some load in the later healing stage to increase the bone density, whereas the implant shields the fracture bone from receiving load. Implants made of bio-resorbable composite materials provide an opportunity to match the properties of bone whilst transferring load to the healing bone more appropriately, along with being easily formed and their degradation products should be tolerated by the human body. The ideal replacement for traditional metallic bone fixation devices should have excellent biocompatibility, be fully bio-resorbable and have sufficient mechanical properties to support the bone during the early healing stages, before gradually degrading over time. A composite comprising polyactide (PLA) matrix reinforced with phosphate glass fibre (PGF) has been regarded as a desirable replacement of the metallic materials. Although, with the rapid development of the biomedical glass fibre reinforced composite materials, considerable effort has been put into investigating the mechanical properties of these materials, a clear understanding of the mechanism of such materials, namely fatigue, flexural failure etc., are still not obtained. This attributes to a variety of reasons, such as the complexity in the components and structure of a composite material. With the capability of simulating the material at a range of scales with reasonable account of complexity, the numerical methods have been adopted as a powerful tool, along with the laboratory tests, to study the composite materials. In this spirit, therefore, the finite element (FE) method was employed in this work to investigate the flexural properties of a biomedical composite material. A representative volume element (RVE) for the composite material was built in the first stage. A layer of finite thickness was assumed between the fibre and matrix to represent the fibre/matrix interface. The RVE was first validated by comparing the numerical predictions of the longitudinal stiffness of the composite with the experimental results. Then the homogenisation method was adopted to predict other mechanical properties of the composites, e.g. shear moduli and transverse moduli, which were not acquired in the laboratory but were required as the input data in the model to simulate the three-point bending test of the composites in the following stage of the study. Based on the results from the first stage, subsequently, an FE model was built to simulate the three-point bending test conducted to investigate the flexural properties of the material. A strain-based continuum damage model (CDM) was selected as the governing equations of the damage onset and evolution in the fibre and matrix, individually. Three different fibre volume fractions of the composite, $20\%$, $35\%$ and $50\%$, were numerically tested against the data from the laboratory. Good agreements were observed from the comparison, particularly, the group of $35\%$ fibre volume fraction showed excellent compliance. It was also showed by the numerical results that the damage accumulation in the fibre was the main drive of the overall failure of the specimen in the three-point bending. In the last stage of the present thesis, a model incorporating fracture bone and a plate-screw implant was built to investigate the `stress shielding' effect for different materials of the implant. Apart from the bio-resorbable composite, the Ti alloy and stainless steel alloy were selected as the representatives of traditional metal. The von Mises stress at the fracture face was calculated to represent the level of `stress shielding' during the healing period. Over the characterized 8-week healing period, the alleviation of the `stress shielding' effect when using the bio-resorbable composite over that of the two metallic materials was clearly observed.
Date of Award8 Mar 2019
Original languageEnglish
Awarding Institution
  • Univerisity of Nottingham
SupervisorJiawa Lu (Supervisor), Xiaogang Yang (Supervisor), Ifty Ahmed (Supervisor) & Lee Harper (Supervisor)


  • Mechanical behavior
  • bio-resorbable
  • biomedical composite material

Cite this