Abstract: A merged binding modeling method for contact areas of bolted joints was presented in order to improve the accuracy of on-line dynamic monitoring and real-time updating of large-scale complex structures' models.During the modeling processes，the contact field elements were  constructed by separating a part of elements from the two mating substructures at each side of the interfaces.The elements in the contact  regions would meet the isotropic conditions after a equivalent treatment，and a merged binding contact relationship was established among the  elements.The greatest advantage of this modeling method is timesaving because it may effectively avoid increasing the numbers of degrees of  freedom in the joint model.Moreover，this method is more universal with advantages of convenience in parameter identification.In the interest of  the consistent accuracy between the calculation model and the physical model，the effects of additional masses of bolts，nuts and sensors on  modeling accuracy were taken into account together with an optimization and identification of physical characterization parameters of bolted  joints based on multi-objective genetic algorithm.The results show that the first six order theoretical mode shapes are in excellent coherence  with their experimental counterparts，and the absolute values of the relative errors of corresponding natural frequency are less than 10%， which shows the proposed modeling method may effectively solve the dynamic modeling problems of joints in engineering practices.

Key words: bolted joint, merged binding, modal analysis, parameter identification

摘要： 为提升复杂结构体模型在线动态监测与实时更新的精度，提出了一种螺栓结合部接触面域的融合绑定建模方法。建模过程中，由配对子结构在接触面两侧各提供一 部分单元组成接触面域单元，通过对该区域单元进行等效处理，使其满足各向同性条件，而接触面域单元之间则被构建为融合绑定接触连接。该建模方法最大的优点是可以有效避免增加结合部模型求解的自由度数，节约计算资源且参数识别方便，通用性较强。为保证计算模型与物理模型具有协调一致的精度，建模过程中考虑了螺栓螺母及传感器附加质量对建模精度的影响，并引入多目标遗传算法对结合部的物理表征参数进行了优化识别。结果表明：前6阶实验振型和仿真振型完全一致，相应固有频率相对误差的绝对值均在10%以内，说明该建模方法能够较为有效地解决工程实际中结合部的动力学建模问题。

关键词: 螺栓结合部, 融合绑定, 模态分析, 参数识别