Response parameters modeling of abrasive jet machined composite using artificial neural network

Abstract

Machining of polymer based nanocomposite is very challenging. Though these materials possess unique properties and developed for specific applications. Therefore analysis of material processing performance of a particular machining process is become essential. Performance of a process can be understood by establish a mathematical relationship between input and output parameters. This empirical relationship can be useful for election of significant variable factors. Once these factors elected then better quality characteristics can be obtained at economical cost. In present study mathematical models for surface roughness and kerf taper of abrasive jet machined nanocomposite are developed using artificial neural network. To check the accuracy of developed models the predicted values of consider quality characteristics are compared with previous published experimental values. Input parameters jet pressure, traverse rate, weight percentage of multi-walled carbon nanotubes and stand of distance for abrasive jet machining of carbon fiber reinforced polymer composite are considered. Average percentage prediction error found less than 5% that shows predicted values from developed models are accurate.