ALTERNATIVE ARTIFACTS FOR EVALUATING SCANNING CMM PERFORMANCE

Coordinate measuring machines (CMMs) with continuous-contact scanning capabilities are being widely used throughout many industries. Most CMM users, if asked, say they require high density scans at high speeds with high accuracy. Obviously, there will be some point at which the accuracy of the scanned data begin to decrease with an increase in the scanning speed. Previous research [1] at UNC Charlotte has revealed that the scanning errors increased in a predictable relationship with the speed when measuring ring gages. There was also found to be a large difference between the results obtained during single point probing and scanning on the same CMM. Other research [2] at UNC Charlotte has developed a compensation model to reduce the scanning errors at high speeds, but this model is only valid for circular features. To date, our literature search has found no standard method (aside from ISO 10360-4) for comparing different CMMs based on their ability to scan at different speeds. A necessary step in comparing scanning CMMs to each other, or to discrete point CMMs, is to use a standardized test method [3]. The current ISO 10360-4 [4] standard describes circular scans on a calibrated sphere. As this sphere is of good quality, we feel that this certification test is more an exercise of the CMM motion controller than a test of the behavior of the measuring system under the excitation of measuring a real part surface. Other tests [5] have been proposed, including one that involves scanning over a gage pin with the same diameter as the probe stylus. This test produces double contacts between the stylus and the artifact where the pin is first contacted, and resulted in large errors at higher scanning speeds. Our current work examines the roles of a linear and a circular artifact, both with a sinusoidal waveform superimposed on a flat surface. By supplementing the ISO scanning tests with the measurement of these artifacts along different machine axes, and at different speeds, we can excite different frequencies in the probing system and in the machine structure. Our goal is to develop a means of distinguishing between scanning CMMs based on their accuracy at different speeds, direction of scanning. This paper describes our work in developing and measuring artifacts that will better capture how a scanning CMM will perform when measuring actual parts.