情報量の概念を用いた磁気ディスク基板の超精密切削条件の決定

概要

論文の詳細を見る
In order to improve the productivity of magnetic disk substrates, the machining process has been devised in which the finish facing is done simultaneous with the rough facing. In this paper, the cutting parameters for this machining process, such as distance between roughing and finishing tools, depyh of cut, feed rate, and work sprindle speed, are discussed, because it is necessary to produce the disk substrates in conformity to the accuracy standards established by the industries. An approach to the problem is based on the concept of information amount and totally different from the conventional one. In the begining, turning experiments of the disk substrates were carried out under various values of the cutting parameters and then their machining accuracies, such as straightnesses of circumferential and radial profiles, radial deviation, outer dia. bump, and flatness, were investigated. The collected data samples were analyzed by the inductive learning algorithm and the information integration method and the following successful results were obtained. (1)By evaluating the information entropy of the data samples, the rules for the selection of the cutting parameters could be framed. (2)By integrating the information amount of the data samples, the optimal values of the cutting parameters could be determined.
社団法人日本機械学会の論文
2000-03-25