Decision methodology of micro end-milling condition using tool catalog data-mining system

Data-mining methods were applied to support decisions about reasonable micro end-mill cutting conditions (cutting speed, feed rate, axial depth of cut and radius depth of cut). The aim of this research was to excavate new knowledge with the mining effect by applying the data mining process of hierarchical and non-hierarchical clustering methods to micro end-mill tool catalogs. Micro end-mill shape element of catalog data were focused on visually grouped end-mills, which meant the ratio of tool shape dimensions. With these process, principal component analysis was used to quantify the correlation degree between cutting conditions and tool shape parameters. End-milling condition decision equations were derived from response surface method using significant predictor variables consisting of tool shape parameters and workpiece mechanical properties. The catalog-mining system appeared to be effective for mining knowledge hidden in a large amount of catalog data related to tool shape and end-milling conditions. Therefore, it appears to be straightforward for unskilled engineers to visually determine micro end-milling conditions from the tool shape. Moreover, short-delivery manufacturing with less waste may be possible.