【摘要】 为了有效识别工程结构中的螺栓连接松动，防范结构连接失效风险，以应变频响函数（SFRF）作为响应特征指标，提出了一种基于模型修正思想的螺栓连接结构松动识别方法。对含4个螺栓连接的搭接板结构，采用薄层单元描述螺栓连接，分别提取仿真模型和松动损伤结构模型螺栓连接位置的应变频响曲线，拾取频响曲线的特征点并构造目标函数。采用拉丁超立方抽样和径向基函数近似方法，构造了连接参数和目标函数之间的代理模型关系。基于模型修正思想，采用遗传算法对目标函数进行优化，识别模型中连接参数值。通过比较模型的初始连接参数和识别出的参数值，可定量描述结构松动损伤。数值算例结果表明，薄层单元有限元建模方式能较准确地模拟螺栓连接特性，使用代理模型可提高搜索效率，能有效识别螺栓连接结构的松动程度。更多还原

【Abstract】 Based on strain frequency response function(SFRF)model updating,an approach that used to detecting the looseness of bolt connections was proposed. A bolt lapped structure with four bolts was simulated by thin layer elements as bolt connections.The objective function was constructed by extracting feature points on the strain frequency function that calculated by simulation model and damage model. Latin hypercube sampling and radial basis function methodology were adopted to construct the surrogate between connection parameters and objective function. Based on the framework of model updating and genetic algorithm,the connection parameters of damage model were identified,and the degree of damage was quantified by comparing the parameters value in initial model and damage model. The results show that the approach proposed can represent bolt connections accurately and the extent of damage can be identified precisely and efficiently.更多还原