Real-time thermal modelling approach of a machine tool spindle based on bond graph method

• Published in: International journal of advanced manufacturing technology Vol. 113; no. 1-2; pp. 99 - 115
• Main Authors: Yang, Yun, Du, Zhengchun, Feng, Xiaobing, Yang, Jianguo
• Format: Journal Article
• Language: English
• Published: London Springer London 01.03.2021; Springer Nature B.V
• Subjects: CAE) and Design; Computer-Aided Engineering (CAD; Empirical analysis; Engineering; Errors; Heat generation; Industrial and Production Engineering; Industrial applications; Machine tools; Mathematical models; Mechanical Engineering; Media Management; Model accuracy; Modelling; Original Article; Parameter estimation; Real time; Spindles; Thermal analysis; Thermodynamic properties; Workpieces; Real-time thermal model; Bond graph method; Verification; Calibration; Machine tool spindle
• DOI: 10.1007/s00170-021-06611-8
• ISSN: 0268-3768

The thermal characteristics of spindles have a significant influence on the workpiece quality. This paper proposes a novel real-time thermal characteristic modelling approach for spindles based on the bond graph method, with simplified thermal structure and brief estimation and calibration of thermal parameters. First, the thermal characteristics of this spindle are analysed and simply divided into different thermal components. Then, a network of thermal capacitances and thermal resistances is established based on the mechanism analysis of heat generation and heat transfer, and consequently, a thermal characteristics model is developed based on the bond graph method. Parameters of thermal conditions of the spindle are briefly estimated using theoretical analysis and empirical formulas. An experiment is designed to calibrate these thermal parameters of the model, followed by verification experiments of accuracy and robustness, the results of which indicate stable good prediction performance, with the maximum error of 0.7355 °C and the average error of 0.1989 °C. Finally, this model is applied in the quantitative investigation of the influence of working conditions on the thermal characteristics of the spindle and the real-time prediction of spindle’s thermal deformation with 5.26 μm maximum error and 1.45 μm average error. The results indicate that this approach has considerable advantages in the real-time prediction of thermal behaviours of spindles and can be used in industrial applications.