Thermal error modeling of spindle based on the principal component analysis considering temperature-changing process

Thermal error is one of the most significant factors that influence the machining accuracy. Thermal error compensation is an easy and economical way to reduce the thermal error. However, the effectiveness of compensation is highly dependent on the thermal error model. In this paper, the thermal characteristics of the spindle of a horizontal machining center are investigated and tested. The measured temperature changes during a period of time are reconstructed based on PCA and used as the new input variables rather than discrete temperatures collected at some certain moments for axial spindle thermal error modeling. By comparing with the thermal error model built on original temperature data, it finds out that the accuracy of spindle thermal error modeling and prediction are improve by adopting PCA for data processing.