Online Vibration Detection in High-Speed Robotic Milling Process Based on Wavelet Energy Entropy of Acoustic Emission

Abstract

This work used acoustic emission (AE) technique to detect machining vibrations during robotic milling process, and elaborated the impact mechanism of milling vibrations on surface roughness and residual stress. The findings indicated that the features relating to machining vibration included a sudden increase of amplitude in the time domain, and variations of frequency distribution in the frequency domain. The duration of machining vibration was exceedingly brief, and the changes of frequency distribution were mainly concentrated in 150–730 kHz. For the processing of AE signals, wavelet energy entroy (WEE) was selected as a detection indicator to monitor machining vibration. A laser vibrometer was also used to collect radial vibration signals for verification, which have similar characteristics with AE signals, confirming the effectiveness of vibration detecting based on AE method. At low spindle speeds, machining vibration is prone to occur at the cut-in and cut-out positions, and tends to become more frequent with the increase of feed speed. High spindle speed and low feed speed can effectively avoid the occurrence of machining vibration. The severe machining vibration occurred when the milling speed was set at 10,000 rpm with feed speed of 1440 mm/min. The influence of robotic milling vibration on surface integrity was also evaluated in details. The amplitude and frequency of machining vibrations during the robotic milling process are random, making the impact mechanism on surface integrity highly complex. Depending on specific conditions, these vibrations could result in deteriorated milling surfaces.