Correlation study of tool flank wear with machined surface texture in end milling S. Dutta a , A. Kanwat b , S.K. Pal b,⇑ , R. Sen a a CSIR-Central Mechanical Engineering Research Institute, Durgapur, India b Mechanical Engineering Department, Indian Institute of Technology, Kharagpur, India article info Article history: Received 12 December 2012 Received in revised form 6 May 2013 Accepted 12 July 2013 Available online 22 July 2013 Keywords: Tool condition monitoring End milling Texture analysis GLCM Run length statistics abstract Indirect tool condition monitoring technique using surface texture analysis is gaining a parallel improvement with the advances of digital image processing techniques with the advent of high- end machine vision systems for fulfilment of high product quality. In this work, condition monitoring of HSS mills and coated carbide milling inserts has been per- formed by analyzing the resulting end-milled surface images using image texture analyses. The machined surface images were pre-processed by recovering them from inhomoge- neous illumination and then two texture analysis methods, namely, gray level co-occur- rence matrix (GLCM) and run length statistical (RLS) techniques were applied on the pre-processed images. Texture descriptors obtained have been highly correlated with the trend of flank wear. Finally a selection of texture features, namely, contrast and GLN, has been made within those extracted texture features for best correlation with tool wear values. Ó 2013 Elsevier Ltd. All rights reserved. 1. Introduction The trend towards automation in machining has been driven by the need to maintain high product quality with improving production rates. These process improvements are possible by tool condition monitoring (TCM). Main pur- pose of TCM is to avoid machine tool downtime. Excessive wear and breakage of the cutting tool is one of the severe causes of downtime. Two types of TCM are there: direct and indirect TCM. In direct TCM, cutting tool wear (flank wear, crater wear, nose wear, breakage, fracture etc.) are measured directly with the help of tool maker’s micro- scope, optical microscope etc. In case of indirect TCM, a de- gree of tool wear can be evaluated by studying different behavior of machine tools, machining process or machined surface qualities without measuring cutting tool wear. So disassembling of cutting tool is not required in this case. Several sensors such as current, force, acoustic emission, power, and surface profiler are used for indirect tool condi- tion monitoring. Now-a-days research is going on tool con- dition monitoring systems using digital image processing for implementing a non invasive TCM technique [1]. Tool images are captured and processed for automatic measure- ment of tool flank wear and crater wear in direct TCM. Flank wear widths were measured automatically from the captured flank surface images using image pre-pro- cessing, thresholding and morphological operations [2–5]. For simplification of the automatic wear measurement, Kerr et al. [6] did three types of image texture analyses of captured flank wear images of turning and milling tools for progressive tool wear. They utilized gray level co- occurrence method (GLCM), fractal analysis method and frequency domain analysis method and extracted few texture descriptors to describe the progressive wear of cutting tool. Finally, they found that extracted features from GLCM technique had better correlation with tool flank wear. Prasad and Ramamoorthy [7] measured crater wear depth using a pair of stereo images taken from two different positions. Finally a back propagation neural 0263-2241/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.measurement.2013.07.015 ⇑ Corresponding author. Tel.: +91 3222 282996; fax: +91 3222 255303. E-mail address: skpal@mech.iitkgp.ernet.in (S.K. Pal). Measurement 46 (2013) 4249–4260 Contents lists available at ScienceDirect Measurement journal homepage: www.elsevier.com/locate/measurement