Prediction method of thermal errors of the screw system in lathes based on moving thermal network.

The positioning precision of the ball screw system in machine tools is primarily affected by the thermal deformation of the screw. Hence, the adaptive prediction method must be established to calculate thermal errors of the ball screw system. First, based on heat transfer theory, a new adaptive moving thermal network model of the screw system with moving thermal excitation and varying operating conditions is proposed. The time-serial data of kinematic-pair representative temperatures and the time-serial position data of the moving nut are used as inputs. This adaptive model is used to determine the temperatures of the key elements in the supporting bearings and those of the moving nut and key points of the screw shaft. Further, the heat generation rate of each kinematic pair can be obtained simultaneously. Furthermore, a real-time prediction method of thermal errors of the screw system with moving thermal excitation under various operating conditions is presented. The effectiveness and robustness of the real-time prediction method of thermal errors was verified experimentally. This method is suitable for the rapid prediction of thermal errors of screw systems by monitoring the real-time kinematic-pair representative temperatures and real-time position of the moving nut. • A new adaptive moving thermal network model for a screw system is proposed. • Based on the moving network model, the real-time thermal error model is proposed. • The heat generation rates of the heat sources are simultaneously identified in real-time. • The presented model can be used for the multiple varying working conditions. • This model has self-adaptive ability and robustness. [ABSTRACT FROM AUTHOR]