Artificial neural network modeling of the drilling process of self-lubricated aluminum/alumina/graphite hybrid composites synthesized by powder metallurgy technique

Abstract

In recent years, the consumption of metal matrix composites (MMCs) materials in many engineering fields has increased enormously. Most industries are usually looking for replacement of ferrous components with lighter and high strength alloys like Al metal matrix composites. Despite the superior mechanical and thermal properties of particulate metal matrix composites (PMMCs), their poor machinability is the main drawback to their substitution to other metallic parts. Machining is a material removal process which is important for many stages prior to the application or assembling of the components. Accordingly, the need for accurate machining of composites has also increased tremendously. This study addresses the modeling of the machinability of self-lubricated aluminum/alumina/graphite hybrid composites synthesized by powder metallurgy (P/M). In the present work, a feed forward back propagation artificial neural network (ANN) system is used to investigate the influence of some parameters on the thrust force and cutting torque in the drilling processes. Experimental data collected were tested with artificial neural network technique. Multilayer perceptron model has been constructed with feed forward back propagation algorithm using the input parameters of cutting speed, cutting feed, and volume fraction of the reinforced particles. Output parameters were the thrust force and cutting torque. On completion of the experimental test, an ANN is used to validate the results obtained and also to predict the behavior of the system under any condition within its operating range. The predicted thrust force and cutting torque based on the ANN model were found to be in a very good agreement with the unexposed experimental data set. The modeling results confirm the feasibility of the ANN and its good correlation with the experimental results. The degrees of accuracy of the prediction were 93.24% and 94.17% for thrust force and cutting torque, respectively. It is concluded that ANN is an excellent analytical tool, which can be used for other machining processes, if it is well trained.