【摘要】 目前关于框架结构损伤诊断的研究大部分都着眼于框架结构梁柱单元的损伤,而实践证明,地震荷载下框架结构在梁柱连接节点处更容易发生损伤。然而对于复杂结构,其有限元模型存在大量的自由度,进而需要大量的传感器,相应的计算量也会加大。为了解决这一困难,文章首先用静力凝聚的方法把有限元模型降阶,然后再跟扩展卡尔曼滤波损伤识别算法结合起来进行结构的损伤识别,用一个6层钢框架的振动台试验数据对该方法进行试验验证,该试验数据由我国台湾大学罗俊雄教授提供。基于试验数据结果表明,将静力凝聚跟扩展卡尔曼滤波算法结合起来的算法不仅能识别出地震激励下梁柱节点损伤的位置,还能定量的识别出节点损伤的程度。更多还原

【Abstract】 Recent research for damage detection mainly focuses on beams and columns of structures under seismic excitation. While in practical applications,it has shown that the occurrence of damage in the beam-column joints more frequent. For complex structures,a great deal of degree of freedom should be taken into consideration when one establishes a finite element model. Thus,a large amount of accelerometers are needed,and additionally large amount of calculation should be involved. To solve this problem,at first the static condensation technique is brought in to reduce the order of finite element model,then the damage detection can be performed by applying the extended Kalman filter.The capability of the proposed method in detecting the damage of joint connections in structures will be further demonstrated by using a shake table experimental data of a six-story steel frame which is provided by professor ChinHsiung Loh,Taiwan University,China. The experimental results show that the proposed method is capable of identifying not only the damage locations but also the damage severities of joints under seismic excitation.更多还原