Vision measurement system for position-dependent geometric error calibration of five-axis machine tools

The geometric error calibration of rotary axes is a key procedure for improving the machining accuracy of five-axis machine tools. In conventional research, the measurement scheme is mainly based on commercial instruments (e.g., the laser tracker, double-ball bar, and R-test). In this study, a vision-based measurement system and an identification algorithm are proposed to measure three-dimensional (3D) displacement and identify 10 position-dependent geometric errors (PDGEs) in the rotary axis ( A and C axes). First, based on a high-precision microscopic checkerboard, the 3D displacement of a chessboard corner is identified using the depth-of-focus method (in the z direction) and the OpenCV corner detection method (in the x and y directions). Second, a simplified error identification model is proposed to identify the 10 PDGEs. Then, the vision system is established to reduce the interference of other installation errors and separate the PDGEs of the A and C axes. Its measurement performance is compared to the R-test. Experimental case studies show that PDGEs are reduced by approximately ten times after compensation, and it proves the accuracy and feasibility of the vision measurement system.