Study on Generation and Optimization Methodology of On-Machine Measurement Schemes for Multi-Axis CNC Machine Tool

Quality Control (QC) is one of the most important phases in the production process. In tradition, the workpiece must be inspected for dimension and shape errors in the QC lab or Coordinate Measuring Machine (CMM) after being machined, which is time-consuming and labor-intensive, especially for large scale parts. Thus the only practical way to do that is to migrate the critical primary inspection operations upstream from the QC lab or CMM to the production floor, and that’s what On-Machine Measurement (OMM) does. OMM technology can inspect the workpiece on its operating station by motion control and the position detection function of CNC machine tool, by which the measuring error caused by the misalignment between measuring datum and manufacturing datum can be avoided accurately. Therefore, it is a feasible way to achieve the real-time control of the manufacturing process and improve the manufacturing accuracy and efficiency. The servo axes of multi-axis CNC machine tool might be redundant for being used in OMM, which will result in that one measurement task could be done by different servo motion schemes. In addition, the map between the machine accuracy and measuring accuracy has not been found out previously. Thus, how to determine metering schemes to measure a workpiece with high accuracy and efficiency is one of the most important problems and also a hotspot for everybody. To solve this problem, a generation method of task-oriented OMM scheme is proposed. Utilizing multi-body system theory and transformation of homogeneous coordinates, a function relationship between the geometrical characteristic of workpiece as independent variables and motions of machine tool as dependent variables is established. All possible metering schemes are obtained by analyzing the solution of functional equation. Then the performance of measurement scheme can be evaluated by comparing the measuring errors of each scheme after building a mapping relationship between the machine tool volumetric errors and measuring errors. In the end, a case study was accomplished, and the correctness and efficiency of the methodology has been verified.