Title: Prediction of surface roughness of 6061 aluminium alloy end milling: a machine vision approach
Authors: Nathan Dhanapalan; G. Thanigaiyarasu; K. Vani
Addresses: Department of Mechanical Engineering, Anna University Chennai, Tamilnadu, India ' Rajalakshmi Engineering College, Chennai, India ' Department of Information Science and Technology, Anna University Chennai, Tamilnadu, India
Abstract: Surface roughness plays an important role to improve the productivity of the component. Prediction of surface roughness becomes the challenging task in the manufacturing environment. In this work, surface roughness of machined surface of aluminium alloy 6061 is predicted from the image features extracted from images of machined surfaces using machine vision. Machine vision paves a platform to predict the surface roughness of machined surface in non-contact method using CCD camera. Aluminium alloy 6061 is machined using conventional vertical milling machine for various cutting speed, feed and depth of cut combinations. The surface roughness values are determined experimentally using stylus instrument. The artificial neural network (ANN) controller is designed to predict the surface roughness of machined surface from the image features. Image features such as skewness, kurtosis, entropy, mean and standard deviation are given as input parameters for training the neural network and surface roughness value measured experimentally have been given as target values. A regression between input and target parameters has been achieved using neural network to predict the surface roughness of the machined surface.
Keywords: artificial neural networks; ANNs; response surface methodology; RSM; analysis of variance; ANOVA; multilayer perceptron; orthogonal arrays; machining; surface roughness; surface quality; aluminium alloys; end milling; machine vision; image features; feature extraction; CCD cameras; cutting speed; feed; depth of cut; skewness; kurtosis; entropy; mean deviation; standard deviation.
International Journal of Machining and Machinability of Materials, 2014 Vol.16 No.3/4, pp.285 - 302
Received: 22 Dec 2013
Accepted: 08 Jun 2014
Published online: 14 Feb 2015 *