The step-diagonal test modifies the diagonal displacement test for machine tool performance evaluation by executing a diagonal as a sequence of single-axis motions. The proposed modification has gained interest because of the claim that the obtained additional data enables the estimation of parametric machine errors, in particular straightness and positioning errors. This paper addresses the properties of the estimated errors for machines with significant angular errors. The theoretical analysis shows that angular errors can cause significant errors in the obtained straightness and positioning errors if their effects are ignored in the analysis. Use of a retro-reflector eliminates the thus introduced bias for some angular errors. The effects of some angular errors can be corrected by repeating a step-diagonal with a different sequence of axis motions. The thus estimated angular errors are expected to have larger uncertainties than those obtained through direct measurement. Our analysis confirms that setup errors in the alignment of the return mirror cause significant errors in the slope of the estimated positioning errors that cannot be detected from the results of the (step-) diagonal test. Correction requires information on the slope of the positioning errors of two axes. Finally, it is not possible to uniquely allocate non-repeatable machine behavior to the various parametric errors.
Proceedings Title: Laser Metrology and Machine Peformance VII, 7th International Conference and Exhibition on Laser Metrology, Machine Tool, CMM & Robotic Performance, LAMDAMAP 2005
Conference Dates: June 24-July 1, 2005
Conference Location: Bedford, UK
Conference Title: 7th International Conference and Exhibition on Laser Metrology, Machine Tool, CMM & Robotic Performance, LAMDAMAP 2005
Pub Type: Conferences
machine tool, machine tool metrology, performance evaluation, step-diagonal test