Geometric error compensation method using the Laser R-test

Abstract

Abstract Regarding the precision of the five-axis machine tool, it is found that 70% of its error is derived from the rotational axis PIGEs and PDGEs. At present, the measurement of these errors is required to be conducted by instruments such as interferometers, DBB, IBS, etc., which are considered expensive and time-consuming. In this research, the development of a contactless calibration tool system, which is designed to measure all 43 items of PIGEs and PDGEs in the five-axis machine tool, is pursued. The calibration tool, named “Laser R-Test” and developed since 2000, is recognized as being capable of analyzing the eccentricity error (XOC,YOC,ZOA,YOA) and angular position error (ECC,EAA). Moreover, the five-axis simultaneous trajectory error can be measured. Through extensive research, it has been determined that the total error in a five-axis machine tool can be controlled to below 40 um after compensating for two error parameters. However, achieving a total error below \(\pm\)10 um remains elusive. This limitation is attributed to the presence of wobble error (such as AOC, BOC, AOB, COB) in the rotary axis orientation, and the analysis and compensation for the rotary axis have yet to be achieved. In this study, a methodology based on Laser R-test and Rodrigues' rotation formula was developed to establish a PIGEs error model. The path XYZ errors were obtained through a simplified path combined with the Laser R-Test. By having the XYZ errors from each point in the path inputted into the model, the current errors and compensation values of the machine tool were determined. The analysis and compensation of 8 PIGEs errors in the rotary axis were completed in 30 minutes using the inspection path. Ultimately, the system's compatibility with K1, K2, and K4 path tests, as prescribed by ISO-10791-6, was ensured, validating the compensation effects on various parameters. A precision below \(\pm\)10 um was achieved, and the inspection time was reduced by over 50%. Then, by integrating the results of previous related studies, this system and method can complete multiple error detections, including angular positioning errors [13], eccentricity errors [11], and wobble errors, by simply using different paths. This greatly reduces the setup time for future users, enhancing its commercial applicability.