The analysis of vibrations generated during conventional turning is one of the reliable means of determining the condition of sensitive components. In this paper, it is clearly presented the necessary steps to extract the indicators in order to be able to detect tool wear during the main three phases of use (running-in, stabilized wear, accelerated wear). It is clearly demonstrated that there is a relationship between the evolution of wear and the measured quantities (vibrations) during machining. To achieve this, we have carried out several measurement campaigns using metal carbide plate tools. Thus, the vibrations generated during the machining operations were recorded along a single axis on the machine tool using a single axial accelerometer positioned on the turret in the vertical direction, perpendicular to the cutting force. The processing of these signals in both the time and frequency domain has proven that vibrations can indeed be used to detect the level of wear.
In this work, it is a question of the cutting tool wear monitoring in mechanical turning. We did this monitoring in three phases which correspond to the life of our tool. To achieve this objective of improving monitoring, we have used a processing method (EMD) that breaks down a large signal into small signals (IMFs). The cut up or processed signals are yet applied in the temporal (RMS) and frequency (Spectrum) indicators in order to monitor the evolution of the tool in relation to its degradation and to check the reliability of the indicators. The obtained results will be optimized in an on-line monitoring system and incorporated into a microcontroller dealing with its three phases, in order to make the comparison of informations each time they are generated by the machine.
Show abstract
Full Text
