多边界寻优的电主轴热特性数字孪生

Digital Twin for Thermal Characteristics of Motorized Spindle Based on Multi Boundary Optimization

摘要: 针对电主轴热特性有限元仿真误差大，热边界条件无法精确辨识问题，提出数字孪生驱动的电主轴热特性在线监测方法。搭建基于C++和Qt的电主轴热特性数字孪生系统，建立基于最小二乘优化的接触热阻与对流换热系数目标函数，根据关键测温点的实测与仿真温度实时优化热边界条件并输出至ANSYS Parametric Design Language (APDL)，通过后台调用ANSYS完成电主轴热特性分析及数据处理，实现物理实体与虚拟模型实时映射的电主轴热特性数字孪生。实验结果表明，该方法有效提高电主轴热特性分析精度，温度场预测精度达98%，热变形预测精度达96%，为电主轴热优化设计及热误差控制提供依据。

Abstract: In order to improve the finite element simulation accuracy of thermal characteristics of motorized spindle and the identification accuracy of thermal boundary conditions, an online monitoring method of thermal characteristics of motorized spindle driven by digital twin is proposed in this study. A digital twin system for thermal characteristics of motorized spindle is built using C++ and Qt. The objective functions of thermal contact resistance and convective heat transfer coefficient are established based on least square optimization method. The thermal boundary conditions are real- timely optimized using the measured and simulated temperatures of thermal key points and output to ANSYS Parametric Design Language (APDL). The thermal characteristic analysis and data processing of the motorized spindle are completed by calling ANSYS in the background. The digital twin for thermal characteristics of the motorized spindle is realized through real-time mapping of physical entities and virtual models. The experimental results show that the simulation accuracy of thermal characteristics of the motorized spindle is effectively improved using the proposed digital twin-driven online monitoring method, the prediction accuracy of temperature field reaches 98%, and the prediction accuracy of thermal deformation reaches 96%, which provides a basis for thermal optimization design and thermal error control of the motorized spindle.