基于柔度曲率梯度的刚架拱桥损伤识别  

Damage Detection of Rigid-Frame Arch Bridge Based on Flexibility Curvature Gradient

作　　者：周宇[1] 狄生奎[1,2] 项长生[1,2] 王立宪[1,2] 

机构地区：[1]兰州理工大学甘肃省土木工程防灾减灾重点实验室,甘肃兰州730050 [2]西部土木工程防灾减灾教育部工程研究中心,甘肃兰州730050

出　　处：《应用基础与工程科学学报》2017年第6期1211-1221,共11页Journal of Basic Science and Engineering

基　　金：国家自然科学基金项目(51568041);甘肃省自然科学基金项目(2014GS02269)

摘　　要：柔度类损伤识别指标对低阶模态敏感,提出柔度曲率梯度的损伤识别指标并应用于刚架拱桥损伤识别.新指标是通过将柔度曲率矩阵在标量场中进行延拓,求取方阵标量场中心元素的梯度向量而得到.采用实际刚架拱片尺寸建立有限元模型,使用频率相对误差与模态置信准则修正结构动力模型,基于动力修正模型验证了柔度曲率梯度的损伤识别效果,并采用截尾高斯分布的误差模型验证新指标的抗误差干扰能力.研究结果表明:基于动力修正模型的损伤指标检验方法新颖、合理,截尾高斯分布模型可以更好地反映测试误差分布,柔度曲率梯度可有效识别刚架拱片单处与多处损伤,在误差水平10%时仍具较好的识别精度.基于此,本文工作可为柔度曲率梯度在较复杂结构与实际桥梁损伤识别的应用奠定基础.The indicator based on the flexibility was vulnerable to low order modes in damage identification. The new damage detection indicator that called Flexibility Curvature Gradient（FCG） based on the penetrating properties of flexibility matrix was proposed.The new indicator is the gradient vector of the center element in matrix scalar field which was calculated by the extension of flexibility curvature matrix in a scalar field.It was proved in rigid-frame arch bridge model which was established by the measured structural parameters and the dynamic model updated with Relative Error（ER） and Modal Assurance Criteria（MAC） in order to get more real and credible proving results. The capability of noise immunity of FCG was verified by the updating error model based on the model of truncated Gaussian distribution.Results are showed that,the verification method based on the dynamic model updating is original and reasonable,truncated Gaussian distribution error can reflect the test error distribution better,FCG can easily identify the single and multiple damages of numerical example,FCG is workable even when the truncated Gaussian distribution error level has reached 10%.Based on above analysis,the work can lay the foundation for the FCG in damage detecting of complex structure and the actual bridge.

关 键 词：刚架拱桥 动力修正 柔度矩阵 曲率梯度 截尾分布 损伤识别 

分 类 号：TU317[建筑科学—结构工程]