Thermal error measurement and modelling techniques for the five-degree-of-freedom spindle drifts in computer numerically controlled machine tools

Thermally induced errors have been significant factors affecting machine tool accuracy. In this paper, the spindle thermal error is examined, where thermal error components with five degrees of freedom (DOFs) are considered. An effective measurement system has been devised for five-DOF thermal errors, consisting of gap sensors and thermocouples around the microcomputer-interfaced environment. Several thermal error modelling techniques are also implemented for the thermal error prediction: multiple linear regression, neural network, system identification methods, etc. The performance of the thermal error modelling techniques is evaluated and compared, giving the system identification method as the optimum model having the least deviation. The developed system for the thermal error measurement and modelling has been practically applied to a computer numerically controlled machining centre, and the spindle thermal errors are effectively compensated using the microcomputer-machine tool interfaced networks. The machine tool accuracy has been improved about four to five times typically.