首页|基于跨模型模态应变能变化的结构损伤识别两阶段法

基于跨模型模态应变能变化的结构损伤识别两阶段法

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
为提高结构损伤识别精度,采用跨模型模态应变能变化(Cross-model Modal Strain Energy Changes,CMSEC)代替传统的模态应变能变化(Modal Strain Energy Changes,MSEC)建立先定位后定量的结构损伤识别两阶段法.与传统模态应变能(Modal Strain Energy,MSE)计算时仅利用结构损伤后的振型不同,跨模型模态应变能(CMSE)的计算同时利用结构损伤前后的振型.先利用CMSEC构建CMSECI(Cross-model Modal Strain Energy Index)损伤指标来确定损伤位置,筛选出疑似损伤构件后,再利用CMSEC建立定量识别损伤程度的灵敏度方程.采用基于奇异值截断的迭代法求解该方程,以避免迭代过程中可能出现的病态方程问题.根据公式构成可以定性判断CMSEC比MSEC更难被噪声干扰.某简支梁算例表明,本文方法在损伤定位和损伤定量两个阶段均比传统方法具有更高的识别精度.为了深入考察本文方法的优越性,采用蒙特卡洛法从宏观的相对误差上界和具体的相对误差分布这两个视角解释了本文方法优于传统方法的原因.
Two-Stage Structural Damage Detection Method Based on Cross-Model Modal Strain Energy Changes
In order to enhance the precision of structural damage detection,a two-stage approach consisting of first-localization and second-quantification was proposed,which utilizes the cross-model modal strain energy changes(CMSEC)instead of the conventional modal strain energy changes(MSEC).In contrast to the conventional approach of calculating modal strain energy(MSE)solely based on the damaged mode shapes,in the cross-model modal strain energy(CMSE)method,the energy considering both the pre-damage and post-damage mode shapes is computed simultaneously.Initially,CMSEC was utilized to develop a damage indicator,known as CMSECI,for identifying the damage locations.Once the potential damage elements are selected,CMSEC was then used to establish the sensitivity equations needed for damage quantification.To circumvent any ill-posed equation issues that may arise during the iterative process,a singular-value truncation-based strategy was implemented to solve the equations.By analyzing the mathematical equations,it can be concluded that CMSEC is less susceptible to noise compared to MSEC.The example of a simply supported beam demonstrate that the two-stage CMSEC approach is more effective in accurately detecting and quantifying the structural damage compared to the conventional method.To further demonstrate the superiority of the proposed approach,the Monte Carlo method was employed to provide a macroscopic view of the relative error upper bounds and a detailed examination of relative error distributions.

damage detectionmodal strain energystructural health monitoringmode shape expansiondamage sensitivity

朱学坤、李翠、俞记生、张雅儒、伍晓顺

展开 >

江西交通职业技术学院路桥工程学院,江西南昌 330013

江西交通职业技术学院信息工程学院,江西南昌 330013

中交长江咨询集团有限公司,江西南昌 330013

江西理工大学土木与测绘工程学院(南昌),江西南昌 330013

展开 >

损伤识别 模态应变能 模态识别 结构健康监测 损伤灵敏度

国家自然科学基金江西省自然科学基金江西省教育厅科技重点课题江西省交通厅科技课题江西省交通厅科技课题

5186802620202BAB204028GJJ2146012022H00212023H0039

2024

10.15959/j.cnki.0254-0053.2024.01.018

力学季刊
上海市力学会 中国力学学会 同济大学 上海交通大学

力学季刊

CSTPCD北大核心
影响因子：0.289
ISSN：0254-0053
年,卷(期)：2024.45(1)
  • 23