Experimental evaluation on abrasive wear properties of Alstonia macrophylla fiber reinforced polypropylene composite using ANN model

Abstract

Development of natural fiber reinforced polymer composites over synthetic based composites offers numerous benefits such as biodegradability, sustainability, light weight, higher strength to weight ratio and yet cost effectiveness. Most of the components like bearings, pulleys, gears, bushes, pulley linings, brake pads, friction plates etc., are mainly made of composite materials which requires understanding of their tribological behavior. Thus, for the first time, study is undertaken to investigate the two-body abrasion wear behavior of the Alstonia macrophylla fiber reinforced Polypropylene (A-PP) composites using pin on disk machine for the selected input variables which is in accordance with ASTM G 99 standards. A-PP composites consisting 40% of fiber volume are prepared by adopting film stacking method and cured under compression molding machine. Sliding distance, grit size of abrasive paper, and applied load are selected as an input variable and their output responses measured are coefficient of friction (COF) and specific wear rate (SWR). Experimental results revealed that, both COF and SWR are found to decrease with increase in the applied load. Moreover, finer abrasive paper (1000 grit size) found to be less effective against the A-PP specimen compared to coarse abrasive paper (320 grit size). Formation of thin polymer film and accumulation of wear debris between the grits during the increase of load are found to be more appealing reasons for decreasing COF and SWR. Same observations are exposed by the FESEM analysis. Furthermore, feedforward backpropagation artificial neural network (ANN) is implemented to predict the aforesaid output responses. Confirmatory results showed that experimental and predicted values are within the acceptable range. For COF, error in prediction within ±6%, whereas for SWR it is ±7%.