Deadline for manuscript submissions: 31 March 2026.
The ultrasonic machining product depends heavily on tool shape, as the tool replicates the workpiece surface. Because of this, tool wear during ultrasonic machining (USM) needs in-depth analysis, as it changes tool geometry and can affect the dimensions of the machined cavity. This study examines an image-based reverse-engineering workflow for characterizing changes in USM tool geometry over cumulative machining time. Images captured using a digital microscope were processed to obtain three-dimensional point-cloud data, from which tool height, width, and volume were evaluated. Linear Regression (LR), Random Forest Regression (RFR), and Extreme Gradient Boosting (XGBoost) were used to estimate the measured geometric responses. The recorded measurements showed an overall reduction in tool dimensions and volume during the experimental sequence. The model estimates followed the broad trends in the data, although their agreement varied by response and machining stage. The work demonstrates the use of microscope-based image capture and point-cloud processing for tracking tool geometry and provides an initial comparison of regression approaches for the recorded sequence.
