Thermal Error Compensation
Energy Efficiency of Thermally Compensated Machine Tools (BFE)
Thermal errors can cause up to 75% of all deviations on a machined workpiece and are therefore one of the most pressing issues facing high precision production. Thermal errors are the time dependent change of the geometrical errors of the cinematic chain, largely caused by thermal expansion and contraction of the machine structure itself. Thermal error compensation is a pressing issue for the machine tool industry and can achieve significant energy saving potential as the current state requires growing cooling systems a resource instead of an intelligence based approach to mitigate the effects of thermal errors on produced parts. Thermal adaptive learning compensation (TALC) can be used to develop a self-learning compensation model based on on-machine measurement data and other inputs that are easily available such as motor speeds and temperature measurements
Generating these compensation models requires large and representative training data of the entire operation range, which is costly to generate or not always available. Using available CAD information for FEM simulations virtual measurements can be generated in order to train robust compensation models on a wider range of load cases that cannot be achieved purely experimentally. However some experiments are still useful to reduce the simulation-reality gap and calibrate the compensation model to the specific machine tool using transfer learning.
Contact: Sebastian Lang