Monitoramento do torneamento do aço abnt d6 com ferramentas de pcbn

Abstract

The monitoring of production processes is of great industrial interest worldwide. In machining its importance lies in the possibility of optimizing the life of the cutting tool and preventing its breakadge, in addition to controlling the surface roughness of the workpiece, so that the search for improvements in quality and productivity, with reduced costs are constantly increasing. A critical point for the unattended machining process is to identify the exact moment of change the cutting tool, considering its maximum use and respecting the quality limits of the workpiece. With the constant increase in manufacturing systems and therefore competitiveness, the need for improvements in the process arises, in which the detection of premature failures that may occur during manufacturing contribute in this sense. With this focus, this work has monitored the turning process of hardened (60 HRC) VC131 (ABNT D6) steel via acoustic emission, main motor electrical current and machining force in order to detect failures, breakage and wear of PCBN cutting tools during tool life tests. This application is justified because of the high cost and low machinability of the work material (ABNT D6 steel) and the high acquisition cost of the cutting tool (PCBN). The effects of the machining parameters (cutting speed and feed rate) on the monitored signals were studied. In addition to monitoring the signals of acoustic emission, electrical current and force during machining, periodically the test was interrupted to measure the flank wear (average - VBB and maximum - VBBmax) and the test was considered complete when the flank wear reached the stipulated end-of-life criteria (0.6 mm for VBBmax). The results of the tests were used to training an artificial neural network in Matlab. An important factor is that this system allows online monitoring, i.e. during normal operation of the machine, without having to interrupt the process during the machining and instant follow up of the cut, allowing prevention of breakage and breakdowns. The force and power signals increased over time, unlike the acoustic emission signal. The neural network presented an excellent feature for detection of wear.