An Innovative Approach to Increase the Accuracy of Multi-Axis Machines for Process-Intermittent Inspection

Abstract

A method for enhancing the effectiveness and robustness of process-intermittent inspection is presented. The method is developed first using mathematical models and measurements closely related to the real parts, thus reducing the uncertainty in both the error estimation and compensation for multi-axis machines. A predictive search algorithm is then developed to identify the minimum number of appropriate measuring points (with respect to the designated tolerance) for arbitrarily shaped manufacturing parts. The identified measuring points are robust in that they least sensitive to uncertainties in measurement and modeling. The search algorithm uses computer simulation with information obtained from previous measurements. Consequently, the measurement effort involved in error modeling has been greatly reduced. The error model derived from the method can be used to correct the process-intermittent probing data for a more accurate assessment of workpiece dimensions. Although the method is designed for general application in multi-axis machines (i.e., machine tools, robots, and coordinate measuring machines), this paper focuses on the specific application of a machining center. Experimental results demonstrate the effectiveness of the error modeling method for accuracy improvement of the machining center for process-intermittent inspection.