Progressive tool flank wear monitoring by applying discrete wavelet transform on turned surface images

• Application discrete wavelet transform on turned surface images.
• Selection of mother wavelet and the decomposition level.
• Extraction of two texture features from surface images to monitor tool flank wear.
• Method for on-machine tool and progressive tool wear monitoring.

In this paper, a method for on-machine tool progressive monitoring of tool flank wear by processing the turned surface images in micro-scale has been proposed. Micro-scale analysis of turned surface has been performed by using discrete wavelet transform. A novel methodology for proper selection of mother wavelets and its decomposition level dependent on the feed rate parameter has also been shown in this research. The selected mother wavelets are utilized to decompose the turned surface images at the chosen decomposition level and two features, namely, GRMS and Energy are extracted as the highly repeatable descriptors of tool flank wear. An exponential correlation of GRMS and Energy values with progressive tool flank wear are found with average coefficient of determination values as 0.953 and 0.957, respectively.