A Digital Twin of a Ball Screw Drive Under Different Conditions: A Case Study About Temperature Drift

The ball screw feed drive system (BSFDS) is a fundamental component of machine tools. Thermal error consti-tutes a primary factor contributing to the inaccuracies observed in models and, hence, the quality defects in the product. In order to decrease thermal error, traditional methods focus on the construction of a thermal error model and compensating algorithm. However, the warm-up phase is usually omitted, which can also have an influence on the behavior of BSFDS. The dynamical behavior of BSFDS in the warm-up phase is influenced by temperature and hence difficult to be described by a linear time invariant model. This research constructs a digital twin (DT) to describe the dynamical behavior of BSFDS in the warm-up phase. The digital twin is identified through the linear parameter variable (LPV) algorithm. To evaluate the algorithm, the constructed digital twins are initially validated on a simulation platform and then on a real-world experimental machine.