Characteristics of Acoustic Emission Signals in Machining Using Diamond Coated Tools

Diamond coated cutting tools have been pursued as a costeffective substitute to brazed polycrystalline diamond (PCD) tools in applications such as machining high-strength and lightweight materials. However, coating delamination has been known as the major failure mode of diamond coated tools, which terminates tool life prematurely. Once delamination failure occurs, the tool substrate often subjects to severe abrasive wear leading to catastrophic tool failures that imparts the part quality and interrupts machining operations. Hence, accurate detections and forecasts of coating delamination events can prevent production loss and assist process planning. In this study, the characteristics of acoustic emission (AE) signals when machining a high-strength aluminum alloy and a composite using diamond coated cutting tools were investigated. The AE signals were analyzed in both time and frequency domains at various machining conditions and different cutting times. It was found that AE root-mean-square values decrease considerably once coating delamination occurs. The results also indicate a correlation between the tool condition and fast Fourier transformation (FFT) spectra of AE raw data. In addition, the machining experiments implied that it may be feasible to use AE signals to monitor the condition of diamond coated tools during machining. INTRODUCTION Acoustic emission (AE) refers to the transient elastic waves generated during the rapid release of energy from a localized source within a solid. For metallic materials, acoustic emissions may be associated with plastic deformation, initiation and propagation of cracks, and friction, etc. Since late 1960s, AE has been widely used to detect cracks in pressure vessels, bridges, hydroelectric dams, and composite laminates, etc. [1]. Among the wide range of non-destructive evaluation techniques for detecting cutting tool conditions, AE has been recognized as a feasible method for in-process tool wear monitoring due to its great sensitivity to tool wear [2]. In machining, AE signals can be easily distinguished from signals associated with machine vibrations and ambient noise because of its high-frequency nature, e.g., from order of 10 kHz to MHz [3]. A considerable amount of efforts has been devoted to exploring the relationship between acoustic emission signals and cutting mechanics. Dornfeld and Kannatey performed orthogonal cutting tests, varied the process parameters and recorded the acoustic emission signals generated [4]. The authors indicated that strong dependence of the AE root-meansquare (RMS) voltages on both the strain rate and the cutting speed was observed. Other studies also claimed that AE-RMS values increase almost proportionally with the cutting speed, while the effects of the depth and width of cut are marginal [5]. Regarding to tool wear effects, Iwata and Morikawi reported that the AE-RMS increased significantly as the carbide tool was severely worn in machining carbon steel [2]. It was further shown that the flank wear progression has a more significant effect than the cutting speed on the RMS values and that the AE-RMS stabilizes after the initial increase that is due to the run-in stage of tool wear. According to Dornfeld and Lan, tool fracture and catastrophic failure followed by the nucleation and growth of cracks inside a tool generate burst AE signals [3]. Dissimilar to