Abstract
Recently, in the field of machine tool, measuring instrument, robot, medical industry, information processing and so on, many minute structures are used for high precision or many functions. In this study, relationships between specification of the sliding surface of a minute part and the coefficient of friction or its standard deviation were investigated in experiment. Photo-fabrication and bonding small steel ball for final treatment on the sliding surface were used. It is concluded from the results that (1) Relationship between specification of the sliding surface on a minute part and the coefficient of friction were declared, (2) Specifications of sliding surface with the desired coefficient of friction were designed by the neural network, and conditions and specifications of sliding surface for the most stable behavior were calculated by this system.
