摘要
提出了一种有限长周期支撑结构损伤概率识别方法。基于特征波导纳原理,分析谐调与存在单一扰乱周期支撑结构的自由波传播规律,建立含单扰乱周期支撑结构的频率特征方程。提出一种无量纲自振频率概念,并运用敏感性分析方法,建立有限长周期支撑结构无量纲自振频率相对于基本周期单元刚度变化敏感性矩阵的一般形式,分析表明所提出无量纲自振频率及其敏感性矩阵与周期支撑结构具体几何材料参数无关,避免敏感性分析对于精确有限元模型的依赖。结合贝叶斯理论与马尔科夫链蒙特卡洛(MCMC)模拟算法,实现基于测量自振频率变化的有限长周期支撑结构单元损伤的概率识别。通过对一含螺栓接头的周期支撑梁实验室模型开展研究,对提出方法的正确性与有效性进行验证。
A probabilistic damage detection methodology for periodically-supported structures of finite-length is developed in this paper.By utilizing the characteristic wave receptance method,the free wave motion of a finite-length periodically-suppor-ted structure without disorder and with a single disorder are both analyzed,and the frequency characteristic equation of periodi-cally-supported structures with a single disorder is developed.In addition,the concept of non-dimensional natural frequency is proposed,and the sensitivity matrix of the non-dimensional natural frequencies with respect to changes of the stiffness of the periodic cells is obtained by using the sensitivity analysis technique.It is found that the sensitivity of the non-dimensional natu-ral frequencies and associated sensitivity matrix is independent of the structural geometric and material parameters,and thus the accurate information about finite element model required for sensitivity analysis is avoided.Furthermore,by employing the Bayesian method together with the MCMC sampling approach,the damage scenarios in periodically-supported structures of fi-nite-length are identified with the damage induced changes of natural frequencies.The feasibility and efficiency of the proposed method is demonstrated by the experimental case studies conducted for a periodically-supported beam with various damage sce-narios.
作者
尹涛
尹孟林
贾晓健
YIN Tao;YIN Meng-Lin;JIA Xiao-jian(School of Civil Engineering, Wuhan University, Wuhan 430072, China)
出处
《振动工程学报》
EI
CSCD
北大核心
2018年第1期91-101,共11页
Journal of Vibration Engineering
基金
国家自然科学基金资助项目(51778506)
关键词
周期支撑结构
概率损伤识别
敏感性分析
特征波导纳
贝叶斯理论
periodically-supported structures
probabilistic damage detection
sensitivity analysis
characteristic receptance
Bayesian theorem
