Vibrational Signal Segmentation Method for Aluminum Drilling Monitoring

Aluminum alloys are widely used materials in the automotive and aerospace industries due to their high strength-to-weight ratio. In these high-precision industries, the quality of drilled holes is critical. A poorly drilled hole can result in rework, scrap, and high rejection rates. This paper proposes a novel method for monitoring and segmenting vibrational signals during the drilling process using cost-effective vibration sensors. The objective is to automatically identify key drilling phases from the vibrational signal, including material removal and drill retraction. After comparing various techniques for signal segmentation, the variability or spread of the data signal was the most significant feature in identifying the material removal period. The proposed method involves preprocessing the data followed by signal segmentation using the sharp changes detection method. Experimental results show that the proposed approach achieved 100% accuracy in identifying cutting signals for both constant and varying spindle speeds. This technique has the potential to enhance real-time process monitoring, optimize drilling accuracy, and improve overall product quality in smart manufacturing systems.