The development of tool condition and cutting quality monitoring in PZT/Si wafer micro milling by Neural Network and AE signal

• Main Authors: Ling-Fang Chang, 張齡方
• Other Authors: Ming-Chyuan Lu
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/05055167633364193711

碩士 === 國立中興大學 === 機械工程學系所 === 102 === 　　Micro-mechanical machining holds the advantage of machining complex feature in micro scale with various types of material and draws much more attention recently due to the increase demand in micro device with various types of material. However, the high tool wear rate in machining brittle material such as silicon and PZT limits its application in mass production. 　　The development of tool wear and surface quality monitoring in machining PZT/Si wafer with the diamond thin film deposited micro mill was focused in this study. The system includes the acoustic emission signal collection, feature selection, and classification models. The class-mean scatter criteria were used in the feature selection, and the back propagation neural network was used for the classifier design. In order to collect the data for the system development and verification, an experiment was conducted on a PMC micro machining center with the diamond film deposited micro mill. In machining, the AE signal and cutting force was collected and analyzed for the delamination of the diamond thin film from tool and tool wear, as well as the quality change on the bottom surface of machined slot. 　　The results show that the close relationship between the AE signal and the tool condition was confirmed, as well as the surface quality of machined slot. With the proper selection of features and number of training samples, higher than 94% classification rate can be obtained for the film delamination and tool wear with the developed system in machining the PZT/Si wafer with the thin film deposited micro tool. Moreover, 100% classification rate can be obtained for the quality monitoring.