An investigation on characterization of bio-composites

Abstract

This research work describes the characterization and fabrication of bio-composites as artificial bone implant materials. The fractured bone can be repaired or replaced by artificial bone materials. Many implant materials has been made in the last three decades made of metals alloys ceramics polymers etc. The main problem with the metallic bone implants is the stress shielding and bone regeneration. Polymeric bone implants may overcome these difficulties. This article presents the processing techniques and experimental results of two different bio-composites viz. Hydroxyapatite (HAp)/High density polyethylene (HDPE) bio-composite by micro-injection molding and compression molding technology and Hydroxyapatite (HAp)/High molecular high density polyethylene (HMHDPE) bio-composite by compression molding technology. The initial materials HAp was synthesized by wet chemical precipitation technique, mixed with HDPE granules at 180 0C in a micro-compounder and also mixed with HMHDPE in a twin screw extruder at 220 0C, followed by micro-injection and compression molding and compression molding. Testing of different mechanical properties like tensile, compressive, flexural, impact and two body abrasion wear has been carried out. The experimental result provided shows a good mechanical behavior. An artificial neural network (ANN) model and a fuzzy logic (FL) model have been presented, for the prediction of abrasion wear amount of the fabricated bio-composites. Two body abrasive wear testing has done as per the full factorial design approach as well as it helps to construct the fuzzy prediction model based on Mamdani method. 75% of the response data used for training and 25% data used for testing of the artificial neural network (ANN) prediction methodology. The predicted results are in good (for testing) arrangement with the experimental data with an absolute average percentage error of 0.3470 and 1.8224 for ANN and 4.4542 and 2.7104 for fuzzy logic. It was found that both the fuzzy and ANN model are able to predict the abrasion wear rate in operating conditions with a very high degree of accuracy.