Wear prediction in an Externally Articulating Knee Prosthesis (EAKP)

Abstract

A novel conceptual knee implant has been introduced recently, which can be described as an Externally Articulating Knee Prosthesis (EAKP). The EAKP has been designed to relieve the symptoms of osteoarthritis (OA) by reducing the load experienced in symptomatic OA knee joints, without sacrificing any knee components and through only a minor surgical procedure. Many factors contribute to knee implant failure: fatigue, wear, crack propagation, osteolysis, and loosening. This thesis focuses on one of the crucial possible failures of an EAKP, that of wear at the contacting surfaces in the articulating regions of the implant. The topic of wear, which is very important in almost all types of knee impant, is particularly important for the EAKP because it has been designed to attach to the knee within a small space which requires that the contacting surfaces of the EAKP are smaller than in conventional knee implants. Experimental procedures to investigate wear in knee implants, for example fullknee simulators involving millions of cycles of loading, are time and cost consuming, with complex parameter set-up. Therefore, this study here will involve a computational analysis of the EAKP, validated through results from experiments. Initial wear predictions were carried out using wear simulations developed using the Abaqus Finite Element (FE) software, of simple Block-on-Ring (BOR) geometries and loading. The wear model incorporates the classical Archard's law iii to predict the wear volume and wear depth based on contact pressure and sliding distance predicted by the Abaqus computational model. BOR physical wear experiments were carried out to produce wear coefficients used as the input in the wear FE models, and again to validate the models. A semi-analytical model of the BOR tests was also used to verify the BOR FE model. This analytical method was developed from an elastic foundation model of the BOR geometry and response. In order to predict wear in the EAKP, a realistic knee model consisting of femur and tibia of a right knee (anatomically accurate, as obtained from a real knee) and all EAKP components was developed using SolidWorks. The complete virtual knee model and EAKP were then exported to Abaqus for wear analysis. The virtual knee model was prescribed with realistic kinematics (from X-ray fluoroscopy data) and loading of the knee joint based on ISO 14243-3. Initial BOR tests at loads of 225N were used to determine a wear coefficient of 4.1E-10 mm3N-1mm-1 for a UHMWPE/metal configuration. Further BOR tests and FE wear simulations at loads of 130N, based on this wear coefficient, predicted wear volumes which agreed to within approximately 5%. The semi-analytical model and FE simulation predictions were within 1-7.8% for wear depths, areas and volumes. iv Using the results from the earlier experiments and simulations, the full knee model with EAKP was analysed. Wear volumes were predicted to be approximately 27.74 mm3 (26.21 mg) in one million cycles with a maximum depth of penetration of 1.08mm. Wear rates for more commonly used implants have been reported to be around 17mg per million cycles which is 35% less than for the EAKP. This increase was as expected, because of the small contacting surfaces of the EAKP, but the wear rates are reasonable and point to survival rates of three years or more. This thesis was concerned with validation of a computational model of the EAKP, and used simple combinations of materials. The use of an optimal combination of materials would be expected to result in better wear rates and survival rates. Comparisons were carried out using wear coefficients used in modelling unicompartmental Oxford mobile bearing and Triathlon PKR fixed bearing implants. The combination used was cobalt chromium and low cross-linked polyethylene in the former and with high cross-linked polyethylene in the latter. Wear volumes predicted with the EAKP are 10.6mm3 (10.18 mg) and 2.28mm3 (2.11 mg), with wear depths of 0.5mm and 0.2mm, respectively. In conclusion, the wear in the EAKP was found to be higher than in more conventional implants. This was expected due to the smaller contact areas in the EAKP, which promote higher contact pressures. With the prediction from BOR wear coefficient, the EAKP should survive for over 3 years in terms of attributes: maximum wear penetration and wear volume. However, with advancements in material technologies, further improvement in the wear-resistance should be possible in the future, leading to longer life for the EAKP