基于模型修正的螺栓结合部虚拟材料参数识别及应用

针对螺栓结合部虚拟材料模型建模及参数识别问题，基于结合部显著影响整体结构动力学性能这一特性，提出基于模型修正的虚拟材料动力学模型参数识别方法。针对参数识别中修正方程的病态问题，根据虚拟材料相关参数及结构各阶模态频率相互之间的影响度，构造修正方程的左右加权函数以减轻其病态程度，并通过仿真算例验证参数识别方法的有效性。探讨平板螺栓连接及哑铃状结构用于虚拟材料参数识别的有效性及抗噪性，加工哑铃结构的实验零件，辨识螺栓结合部虚拟材料模型的参数。基于机床螺栓结构的常用工况，建立虚拟材料模型参数库，并在CKX5680数控机床上验证参数库的有效性。结果表明：采用模型修正技术可以准确地识别无噪声时的虚拟材料参数；采用哑铃结构实验试件在有噪声情况下，螺栓结合部虚拟材料参数识别误差小于8%；采用虚拟材料模型模拟螺栓结合部的建模误差小于5.6%。

Parameter Identification and Application of Virtual Material Model of Bolt Joint Based on Model Updating Method

To solve the modeling and parameter identification problem about the virtual material model of bolt joint, based on the fact that joint significantly affect the dynamic property of the whole system, a parameter identification method based on model updating is proposed. The left and right weighting functions, which constructed by considered the influence of virtual material parameters and structural modal frequency, are proposed to reduce the ill-condition of the modified equation. Accuracy of the proposed method is verified by numerical example. The table bolted connection structure and dumbbell structure is discussed for bolt parameters identification, and the dumbbell test structure is designed to identify the virtual material parameters. Finally, based on the common working conditions of bolt joints on machine tools, the virtual material model parameter database is established, and the validity of the parameter database is verified on the CKX5680 CNC machine tool. The results show that the model updating method can accurately identify virtual material parameters in the absence of noise. In the case of testing noise, the parameter identification error for dumbbell structure is less than 8%. The modeling error of virtual material model for bolt joint is less than 5.6%.