Positioning error of feed axis decouple-separating modeling and compensating research for CNC machine tools

Abstract

In order to establish a high-precision comprehensive position errors compensation model for CNC machine tools, a positioning error decouple-separating method is put forward. Temperature data and positioning errors are measured for 3-axis CNC machine tools, the variation of temperature fields and positioning error has been studied. Based on this regularity, a new concept of error-sensitive degree is defined, and constructs error-sensitive degree matrix by grey correlation algorithm, and optimizes the measurement point. Though the research of the position error variation, the thermal error and geometric basic error is decoupled and modeled each other by multiple linear regression and GM(1, n) algorithm. Then position error comprehensive model is established by linear superposition method. Though the on-line compensation experiment on VXC-560 type machine tools under different working conditions, the result shows that positioning error of x axis in cold-state is reduced from 11.1 μm to 4.5 μm, decreasing range is 59.5%. The positioning error of x axis in thermal state is reduced from 34.9 μm to 8.2 μm, decreasing range is 76.5%, and verifies the new compensation idea of directly driving hardware by a error model of machine tools. It has certain engineering application prospect.