Self-calibration for a hexapod coordinate measuring machine

The Hexapod Coordinate Measuring Machine (HCMM) is composed of six telescoping struts with laser interferometric displacement measurement and a single fixed-length center rod. The laser interferometers measure the length changes of each strut. However, in order to calculate the coordinate of the probe, it is necessary to know the absolute length of each strut and the distance between the three joints. Thus there are nine kinematic parameters to be determined in the calibration process. The nine parameters are the three base lengths and the initial lengths of the six struts.; This research presents a method to find the initial strut lengths and the base joint locations without adopting any external artifacts or measurements. During calibration the measuring probe is moved to a number of different positions within the measuring volume. For each position, the center rod length is assumed to remain constant. The length is calculated based on an initial guess at the machine kinematic parameters. A non-linear least squares problem is then formulated to search for the values of the base locations and initial strut lengths which minimize the variation in the computed center rod length.; A computer simulation of the self calibration process has been implemented. Noise content in each laser interferometric measurement signal was monitored to determine its effect on the quality of the calibration results. It was found that the magnitude of calibration uncertainty is proportional to the magnitude of the noise level. The simulation tool also found that the data collected from some specific motions is insufficient for calibration use. Thus, the observation indices were used to determine the usability of a data set. This research also investigated the influence of the number of position used for calibration and how it related to the uncertainty of the identified parameters. Experiments were performed to estimate the thermal time constants of the HCMM components. Thus, one can determine the time period that the HCMM has to be re-calibrated. In the final of this research, we demonstrated how to evaluate the HCMM volumetric error.