A selection method of key temperature points for thermal error modeling of machine tools featuring multiple heat sources

Abstract

A method that integrates grey relational and thermal sensitivity analyses, and fuzzy c-means clustering, called GTF method, is proposed to select key temperature points for thermal error modeling of machine tools featuring multiple heat sources. A two-dimensional temperature-error index is employed to prevent candidate temperature points with high correlations from being excluded when selecting the temperature points to improve thermal error compensation. To verify the method effectiveness and versatility, prediction accuracies were estimated for a vertical machining center and a floor-type boring machine with multiple heat sources. The root mean square error average reduction rates of the GTF method were approximately 28.0 % and 25.8 % in comparison with the conventional method for the two machine tools, respectively. From the results, it was confirmed that the proposed GTF method ensures accurate thermal predictions for machine tools with multiple heat sources, and is versatile.