OPTIMIZATION OF PREDICTIVE MODELS FOR THE THERMAL BEHAVIOR OF LEADING AND THREADED SPINDLES AND THEIR CORRESPONDING ROLLING BEARINGS USING AI AND BIG DATA

The thermal behavior of mechanical components has a critical impact on the accuracy and reliability of machine tools, especially for components such as leading spindles, threaded spindles, and their rolling bearings. During intensive operation, heat is generated due to friction, internal losses in lubrication, and external influences, leading to thermal deformations, increased clearance, and loss of positioning. Managing and predicting the thermal behavior of these components is crucial for ensuring machining accuracy, particularly in high-precision and CNC machine tools. The development of numerical methods such as thermal FEM simulations and the application of deep learning models and artificial intelligence have enabled a deeper understanding and better prediction of thermal effects. Additionally, a large number of new algorithms allow for the optimization of models, particularly in the domain of boundary conditions in thermal models. The aim of this research is to present methods for optimizing models for predicting the thermal behavior of leading and threaded spindles and their bearings under real operating conditions, based on collected temperature data, operating parameters, and experimental validation.