An Image Analysis Algorithm for Measuring Flank Wear in Coated End-Mills  

作　　者：Vitor F.C.Sousa Jorge Gil Tiago E.F.Silva Abílio M.P.de Jesus Francisco J.G.Silva João Manuel R.S.Tavares 

机构地区：[1]Institute of Science and Innovation in Mechanical and Industrial Engineering,Campus FEUP,Rua Dr.Roberto Frias,Porto,4200-465,Portugal [2]Departament of Mechanical Engineering,Faculty of Engineering,University of Porto,Rua Dr.Roberto Frias,Porto,4200-465,Portugal [3]CIDEM,ISEP,Polytechnic of Porto,Rua Dr.António Bernardino de Almeida,Porto,4249-015,Portugal

出　　处：《Computers, Materials & Continua》2025年第4期177-199,共23页计算机、材料和连续体(英文)

摘　　要：The machining process remains relevant for manufacturing high-quality and high-precision parts,which can be found in industries such as aerospace and aeronautical,with many produced by turning,drilling,and milling processes.Monitoring and analyzing tool wear during these processes is crucial to assess the tool’s life and optimize the tool’s performance under study;as such,standards detail procedures to measure and assess tool wear for various tools.Measuring wear in machining tools can be time-consuming,as the process is usually manual,requiring human interaction and judgment.In the present work,an automated offline flank wear measurement algorithm was developed in Python.The algorithm measures the flank wear of coated end-mills and slot drills from Scanning Electron Microscopy(SEM)images,according to the ISO 8688 standard,following the same wear measurement procedure.SEM images acquired with different magnifications and tools with varying machining parameters were analyzed using the developed algorithm.The flank wear measurements were then compared to the manually obtained,achieving relative errors for the most common magnifications of around 2.5%.Higher magnifications were also tested,yielding a maximum relative error of 13.4%.The algorithm can measure batches of images quickly on an ordinary personal computer,analyzing and measuring a 10-image batch in around 30 s,a process that would require around 30 min when performed manually by a skilled operator.Therefore,it can be a reliable alternative to measuring flank wear on many tools from SEM images,with the possibility of being adjusted for other wear measurements on different kinds of tools and different image types,for example,on images obtained by optical microscopy.

关 键 词：Image processing wear measurement MACHINING 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]