Modeling of Laser assisted machining process using Artificial Neural Network

Abstract

Recently machining with the assistance of laser is the cost effective proven technique to process high temperature alloys in aerospace industries. The challenge in laser assisted machining is to optimize the process parameters to get good surface integrity without changing the metallurgical and mechanical properties. As several input parameters such as feed rate, pulsed frequency, depth of cut, laser power and cutting velocity are involved in LAM, experimental studies are not an effective solution. So the main aim of this work is to develop an Artificial Intelligence model to understand the process mechanics and for the prediction of surface roughness, specific cutting energy and cutting zone temperature during laser assisted machining. ANN model has been developed by considering the topology of the network, selection of algorithm and selection of number of neurons. ANN model is developed by considering Levenberg-Marquardt algorithm and error is minimized by root mean square method. There is a better agreement between experimental and ANN model. The proposed model predicts the surface finish and specific energy with an prediction accuracy of 95.94 and 99.618% respectively during machining of hardened steel.