Optimizing the Tribological Properties of UHMWPE Nanocomposites—An Artificial Intelligence based approach

Abstract

The longevity of the hip implants has been a major issue in recent times due to inadequate material used for implants. Since the metal on polymer implants has issues such as tissue degeneration and osteolysis, the focus of this study is to improve the tribological properties of ultra-high molecular-weight polyethylene (UHMWPE) which has been in use on acetabular cup of hip implants by considering multiple nanoparticles like carbon fibre, carbon nanotubes and graphene as reinforcements. It is extremely difficult and time-consuming through numerous experimental trials to arrive at the optimum material composition of nanoparticles. Therefore, an effort has been made on developing a new polymer nanocomposite by utilizing the artificial intelligence (AI)-based design which includes the techniques, viz. artificial neural network (ANN) and genetic algorithm (GA). The input parameters like weight fraction and the geometry of the different nanoparticles related to the tribological properties were collected from various published literatures, and modelling was done through ANN for the output parameters, viz. coefficient of friction and specific wear rate. Best ANN predictive model was chosen individually for each output parameters on iterating the different hidden nodes. The fundamental correlation between the input and output parameters was investigated through sensitivity analysis. Optimization studies were performed using genetic algorithm (GA) with the best-chosen ANN model as an input to get optimum input variables. Thus, the AI-based approach of designing the UHMWPE nanocomposites shows an enhancement on the tribological properties that pave a way for further experimental trials.