Modern industry trends to smart machining that improves productivity at a low cost. The kernel technology of intelligent manufacturing is the automatic on-machine measurement (OMM). When applying OMM technology to computer numerical control (CNC) machines, in-situ measurement takes place in the machining environment without the need of unloading the tool and the part. However, adverse measurement environment, limitations on the efficiency of data capturing and processing, and diversified measured objects render efficient and accurate OMM very difficult. Holistic solutions are needed to advance OMM technology and therefore many scientific topics are involved. This work primarily focuses on geometric modeling of the on-machine cutting tool measurement and kinematic modeling for the calibration process of both the probe and the machine.
On-machine cutting tool measurement often takes place on a laser tool setter. However, the geometry principles of the gauging mechanisms of laser tool setters are complicated and had not been studied before. This dissertation modeled such a gauging mechanism and presented virtual simulations of the measurement processes on laser tool setters based on geometry principles. The virtual simulations can predict and compensate the measurement errors, allowing for accurate tool setter calibration processes in practical situations.
For cutting tool measurement, the tool length characteristic curve for measurement of round-insert mills is discovered. The derivation of the tool length characteristic curve was carried out by modeling the geometries of tool length measurement processes on a laser tool setter. Based on this characteristic curve, an accurate and efficient approach to measuring lengths of mills with round inserts and bottom cutting edge wear is proposed.
Current techniques for probe calibration and machine calibration assume the impractical situations where either the machine is accurate or the location of the probe is accurately known. To address these drawbacks, the actual kinematic model of a six-axis belt grinding CNC machine with a customized add-on probe is built in this dissertation. Using this model along with a specially designed artifact can facilitate the simultaneous calibration of the probe position and the machine geometry error.