Development of AE Signal Based Tool Breakage Prognosis System in Micro Gun Drilling

Abstract

The AE signals were analyzed in this study for monitoring tool conditions in micro gun drilling to improve the reliability of tool condition monitoring system in micro gun drilling. The AE signals corresponding to the conditions before and during tool breakage were analyzed first based on the time and frequency domain features, followed by the development of a prognosis system for the micro gun drilling process. The system consists of three modules, including a signal transform module for feature generation, a feature selection module for selecting proper features closely related to the monitored event, and a classifier for distinguishing the condition just before tool breakage from the others. To collect signals for the feature analysis and the system development, an experiment was conducted on a Swiss type turning machine. A micro drilling tool with the size of 0.9mm in diameter and 125mm in length was used in drilling the stainless steel. At the same time, an AE sensor was installed on the tool holder to collect the AE signal. During the deep hole drilling, the gun drill was fixed on a tool holder and the stainless workpiece rotated to generate the motion for micro drilling. The results show that the feature of AE signal can be observed to be changed not only during tool breakage, but also before the tool breakage occurring. With properly selecting the features closely related to various tool conditions, the condition before tool breakage can be detected successfully by the developed prognosis system.