Thermal error modelling of NC machine tools based on impulse response model

High accuracy of machine tool is the assurance of top-quality products in machining processes. Among the various errors related to machine tools, thermal errors have a significant effect on machining accuracy and directly determine the machined quality of the finished part with both the surface finish and the geometric shape. So, if we establish thermal error model accurately, the affection of thermal error would be eliminated through compensation. How to get the key temperature measuring points and robustness of the model is the main problem of thermal error modelling using traditional methods. Accordingly, the paper presents a novel method for searching key thermal point by means of infrared imaging system and modelling by system identification based on least square principle. The experiments of thermal error compensation have been conducted on a rotary table of five-axis Milling Center. The results show that the key points are selected simply and effectively and the compensation model has good adaptability, thus its prediction is accurate.