A Technique for Enhancing Machine Tool Accuracy by Transferring the Metrology Reference From the Machine Tool to the Workpiece

摘要

High-speed machining (HSM) has had a large impact on the design and fabrication of aerospace parts and HSM techniques have been used to improve the quality of conventionally machined parts as well. Initially, the trend toward HSM of monolithic parts was focused on small parts, where existing machine tools have sufficient precision to machine the required features. But, as the technology continues to progress, the scale of monolithic parts has continued to grow. However, the growth of such parts has become limited by the inability of existing machines to achieve the tolerances required for assembly due to the long-range accuracy and the thermal environment of most machine tools. Increasing part size without decreasing the tolerances using existing technology requires very large and very accurate machines in a tightly controlled thermal environment. As a result, new techniques are needed to precisely and accurately manufacture large scale monolithic components. Previous work has established the fiducial calibration system (FCS), a technique, which, for the first time provides a method that allows for the accuracy of a coordinate measuring machine (CMM) to be transferred to the shop floor. This paper addresses the range of applicability of the FCS, and provides a method to answer two fundamental questions. First, given a set of machines and fiducials, how much improvement in precision of the finished part can be expected? And second, given a desired precision of the finished part, what machines and fiducials are required? The achievable improvement in precision using the FCS depends on a number of factors including, but not limited to: the type of fiducial, the probing system on the machine and CMM, the time required to make a measurement, and the frequency of measurement. In this paper, the sensitivity of the method to such items is evaluated through an uncertainty analysis, and examples are given indicating how this analysis can be used in a variety of cases.