Development of a method for estimating the amount of displacement of a low-rigidity cutting object by machine learning

(Data set by image of processing conditions and displacement）

Abstract

When estimating the effect of elastic deformation of a workpiece generated during machining on machining errors in machining of low rigidity objects, it is necessary to continuously carry out finite element analysis in the shape in the middle of machining and increase in calculation time becomes a problem. Then, in this study, was attempted to drastically reduce processing time by forming a data set obtained by making processing conditions and displacement distribution information after processing to two-dimensional image information, to develop a method guessing displacement distribution by machine learning. In this paper, we explain the results obtained for the shape of the blade.