Critical Aspects of Practical Implementation of Digital Twins in Machining

Abstract

Process digital twins are developed for predictive checks and optimisation of machining processes and for post-process checks of workpiece quality based on machining data. Although there are significant advantages to applying both of these concepts, practical implementation is difficult and hindered by various barriers. This study presents a use case of mould milling on a medium-sized machining centre supported by predictive and post-process digital twins. The study includes quantification of machining errors as well as possible steps for improving the process in terms of workpiece surface quality and machining productivity. Issues related to incorporating digital twins into the process planning workflow are described and a resultant hybrid digital twin (HDT) concept is proposed. The HDT consists of predictive and post-process parts, both in the minimal viable configuration. The predictive digital twin includes a CNC interpolator for simulating the set-point position values and machining time. The post-process digital twin includes a material removal model based on CNC actual feed drive positions which makes it possible to compare programmed and actual workpiece geometry, enabling quick identification of surface errors. The approach is demonstrated and verified on the mould milling use case.