The modeling method on thermal expansion of CNC lathe headstock in vertical direction based on MOGA

• Published in: International journal of advanced manufacturing technology Vol. 103; no. 9-12; pp. 3629 - 3641
• Main Authors: Hou, Ruisheng, Du, Hongyang, Yan, Zongzhuo, Yu, Wenbin, Tao, Tao, Mei, Xuesong
• Format: Journal Article
• Language: English
• Published: London Springer London 01.08.2019; Springer Nature B.V
• Subjects: Accuracy; Ambient temperature; CAE) and Design; Computer-Aided Engineering (CAD; Engineering; Genetic algorithms; Heat sources; Industrial and Production Engineering; Machine tools; Mechanical Engineering; Media Management; Model accuracy; Multiple objective analysis; Numerical controls; Original Article; Performance tests; Prediction models; Robustness (mathematics); Thermal expansion; MOGA; Thermal error; Temperature distribution; Thermal error compensation
• ISSN: 0268-3768; 1433-3015
• DOI: 10.1007/s00170-019-03728-9

Thermal errors often occur in spindle system due to the changing of ambient temperature and/or inner or outer heat sources of machine tools (MTs), which is the major factor restricting the machine accuracy. Therefore, constructing compensation models with high accuracy and high robustness turn to be a cost-efficient approach for minimizing thermal error and improving machine accuracy. There are two categories of modeling methods for thermal errors so far, namely, physical-based method and empirical-based method. Each modeling method has its own merits and drawbacks. In this paper, a multi-objective genetic algorithm (MOGA) was used to combine the approximative physical model based on thermal expansion mechanism in vertical direction of numerical control (NC) lathe headstock with the accurate empirical model based on experimental data obtained from thermal performance test of spindle system. Consequently, a prediction model of thermal error in vertical direction of headstock with high accuracy and high robustness was obtained. The new model synthesized the advantages of the two types of modeling methods and showed relatively high accuracy and robustness. According to the results of a series of verification experiments on thermal performance of the spindle system under various ambient temperatures and different working conditions, the precision of prediction model has maintained above 74%.