A Study of Vertical Machine-tool Thermal Displacement Measurement and Compensation using Back-propagation Neural Networks

• Main Authors: Yuan-Tai Chang, 張元泰
• Other Authors: Shao-Hsien Chen
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/92176954981973177979

碩士 === 國立勤益科技大學 === 機械工程系 === 103 === There are many kinds of reasons affect precision of machine tools, including servo control stability, mechanical deformation, processing load caused by deformation and thermal deformation. The thermal deformation of the total error of more than 50%. Thermal deformation can be divided into internal and external heat source. The spindle system is maximum heat on the machine. Therefore, this study experiments were planned for the ambient temperature, the maximum spindle speed and variable speed operation at no-load tests. This test analyzes the impact changes in ambient temperature and spindle temperature rise of machine accuracy . In addition to explore oil cooler to control the impact of machine accuracy on the spindle temperature ris steady state. The analyzed through the captured data to establish the relationship between the structure of the machine and the amount of deformation temperature. The application back-propagation neural network to predict thermal error. Final validation by actual cutting Internet forecast accuracy. Its experimental measurements showed that the maximum thermal error 12000rpm spindle speed can be controlled at ±5μm. The variable speed spindle machining sequence can improve the thermal error to within 7μm. The random variable spindle speed cutting can improve thermal error can be to within 11μm.