Modern Digital Twin for Validation and Generation of Datasets for Machine Tool Spindle Modeling

The manufacturing industry requires new services to manage machine tools effectively. However, the lack of training data challenges the creation of accurate machine learning or optimization models. To address this issue, this paper proposes a framework based on a Digital Twin that simulates the phenomena of a machine tool spindle. This framework enables the generation of datasets for machine learning by incorporating disturbances, such as vibrations on bearings, from multiple perspectives. The proposed methodology has been validated through a confirmed case study of a machine tool a hardware-in-the-loop real-time simulator.