基于贝叶斯方法的砌体材料损伤识别

VIP免费

3.0 侯斌 2024-11-19 4 4 3.29MB 87 页 15积分

侵权投诉

摘要

砌体结构是一种历史悠久的建筑结构形式。在长期的使用过程中，环境侵蚀、

自身老化、自然灾害因素等联合作用，将导致其材料的损伤累积、构件和结构承

载能力下降。通过无损的检测方法获取动力特性，进而识别材料的损伤情况是当

前研究的热点。

弹性模量是砌体结构安全性的重要指标，也是砌体结构损伤识别研究及有限

元分析的基础参数。为通过无损方法获得此重要参数，首先通过对砌体试件进行

环境激励与人工激励的动力测试，获取砌体试件的基本频率，并验证环境激励识

别砌体试件基本频率的可行性。在此基础上，通过采用马尔科夫蒙特卡洛（ Markov

Chain Monte Carlo, MCMC）抽样的贝叶斯方法，获得砌体试件基本频率的后验概

率密度分布，作为基本频率估计的依据。然后，根据有限元中弹性模量与基本频

率的关系，反算环境激励下砌体试件的弹性模量，并与静力实验弹性模量进行对

比验证。

以环境激励下6个无损砌体试件的弹性模量均值，作为损伤有限元模型的输

入弹性模量，计算有限元模型的初始频率。利用塑料泡沫模拟砌体试件不同位置

砖与砂浆的损伤，根据应变等价原理推导损伤因子与基本频率的关系式，计算损

伤砌体试件的损伤因子。以此损伤因子折减有损砌体试件有限元模型的弹性模量，

对有限元模型进行划分，分别获取不同划分下有限元模型模态分析的频率；基于

MCMC 抽样的贝叶斯方法得到频率的后验概率密度分布，与损伤砌体试件实测频

率的概率密度分布对比，识别砌体试件的损伤位置，并将该方法用于历史建筑的

损伤识别研究。

对一百年历史建筑沿其东西向与南北向进行多次基于环境激励的动力测试，

运用傅里叶变换（Fast Fourier Transform, FFT）与窗函数处理动力测试加速度时程

数据，获取历史建筑的频率数据，通过MCMC 抽样得到频率的后验概率密度分布。

结合有限元模型计算历史建筑的损伤因子，通过比较损伤因子与频率的变化率，

判断历史建筑是否有损伤产生。以前三次测试频率的后验概率密度分布均值作为

历史建筑频率的初始状态，实时监测历史建筑的动态。该方法识别得到的结果具

有唯一性，可以通过多次定时动力测试，实现以经济、有效的方式对历史建筑进

行长期监测。

关键词：贝叶斯方法环境激励基本频率损伤识别历史建筑

ABSTRACT

Masonry structure is one of the structural forms of buildings with a time-honored

history. The combined effects such as environmental erosion, self-aging and factor of

natural disasters result in the materials’ damage accumulation and decline of carrying

ability of structural member over the long period being used. Getting dynamic

characteristics through nondestructive detection method, and then identify the damage

of materials is a hot research at present.

Modulus of elasticity is not only an important indicator of the safety of masonry

structures, but also the basic parameters to identify the damage of masonry structures

and the finite element analysis. To obtain this important parameter by nondestructive

method, first, the dynamic test should be made on the masonry specimens under the

environmental incentives and artificial incentives. Then the basic frequency of masonry

specimens can be obtained, and the feasibility of identification basic frequency with

environmental incentives will be verified. On this basis, the posterior probability density

distribution of basic frequency of masonry specimens can be obtained based on the

Markov chain Monte Carlo (MCMC) sampling Bayesian methods, as the basis of the

basic frequency estimation. After that, the elastic modulus of masonry specimen can be

calculated depending on the relationship of the elastic modulus of finite element and the

basic frequency, then compare the result with static test.

The mean elastic modulus of the six masonry specimens under environmental

incentives can be regarded as the initial modulus of elasticity of damage finite

element model, and then calculate the basic frequency of the damage finite element

model. The damage of brick and mortar of the masonry specimens can be simulated

by plastic foam. According to the relationship between damage factor and basic

frequency derived by equivalent strain principle, the damage factor of damage

masonry specimen can be calculated. Using the damage factor discount the elastic

modulus of damage masonry specimens and divide the finite element model, the basic

frequency under different finite element model can be obtained with the modal

analysis. Then the posterior probability density distribution of the basic frequency can

be gained by MCMC sampling based on Bayesian method. Comparing with the

probability density distribution of measured frequency, the location of masonry

specimen can be identified, and the method can be used to identify the damage of

historic buildings.

The dynamic test based on ambient excitation can be carried many times on a

century-old building along the east-west and north-south direction. Using the Fast

Fourier Transform (FFT) and the window function to process the testing data, the

frequency of historical building can be obtained, and then calculate the probability

density distribution of frequency by MCMC sampling. The damage factor of

historical building can be calculated through combining the finite element, and then

compare the rate of change of damage factor and frequency to determine whether

there is damage on the historical buildings. The mean of the posterior probability

density distribution of three previous testing frequencies can be regarded as the initial

frequency of historical building, and monitor historical building real-time. The results

obtained by this method are unique, and also can be conducted repeatedly, to achieve

long-term monitoring of historical buildings in an economic and efficient way.

Key Word: Bayesian method, ambient excitation, basic frequency,

damage identification, historical building

中文摘要

ABSTRACT

第一章绪论................................................................................................................1

1.1 引言...................................................................................................................1

1.2 结构损伤识别 ...................................................................................................2

1.2.1 结构动力测试方法 ....................................................................................2

1.2.2 测试数据处理............................................................................................3

1.2.3 基于贝叶斯方法的损伤识别 ....................................................................4

1.3 本文研究内容 ...................................................................................................5

第二章砌体材料的力学性能实验 ............................................................................6

2.1 砖抗压强度实验数据 .......................................................................................6

2.1.1 砖试件制备方法 ........................................................................................6

2.1.2 实验结果评定............................................................................................7

2.2 砂浆立方体抗压强度实验 ...............................................................................9

2.2.1 砂浆立方体试件制作 ................................................................................9

2.2.2 砂浆立方体实验结果评定......................................................................10

2.3 砌体试件抗压强度实验数据 ..........................................................................11

2.3.1 砌体试件制作 ...........................................................................................11

2.3.2 砌体试件静力实验 ..................................................................................12

2.3.3 砌体试件弹性模量计算 ..........................................................................14

第三章基于环境激励的砌体试件动力测试 ..........................................................18

3.1 砌体试件有限元分析.....................................................................................18

3.1.1 砌体试件有限元模型 ..............................................................................18

3.1.2 有限元模型校准......................................................................................18

3.2 砌体试件的动力测试 .....................................................................................21

3.2.1 动力测试设备 ..........................................................................................21

3.2.2 测点布置 ..................................................................................................21

3.2.3 动力测试信号采集..................................................................................23

3.3 动力测试数据处理.........................................................................................24

3.3.1 傅里叶变换与窗函数..............................................................................24

3.3.2 动力测试数据处理..................................................................................27

3.4 砌体试件基本频率推断.................................................................................31

3.4.1 贝叶斯推断的实现过程 ..........................................................................31

3.4.2 基本频率的后验概率密度分布..............................................................32

第四章基于贝叶斯方法的砌体试件损伤识别 ......................................................37

4.1 基于环境激励的砌体弹性模量识别 .............................................................37

4.1.1 无损砌体试件弹性模量识别 ..................................................................37

4.1.2 有损砌体等效弹性模量识别 ..................................................................40

4.2 损伤砌体试件损伤因子计算 .........................................................................44

4.3 砌体试件损伤识别 .........................................................................................46

第五章历史建筑频率识别及监测..........................................................................53

5.1 工程概况 .........................................................................................................53

5.2 有限元模型 .....................................................................................................55

5.3 历史建筑动力测试 .........................................................................................58

5.3.1 测点布置 ..................................................................................................58

5.3.2 动力测试设备 ..........................................................................................59

5.3.3 动力测试数据处理..................................................................................59

5.4 历史建筑损伤识别 .........................................................................................64

5.4.1 频率的后验概率密度分布......................................................................64

5.4.2 历史建筑损伤识别 ..................................................................................67

5.4.3 历史建筑健康监测..................................................................................68

第六章结论与展望 ..................................................................................................71

6.1 结论.................................................................................................................71

6.2 展望 .................................................................................................................71

附图 ............................................................................................................................73

参考文献....................................................................................................................80

在读期间公开发表的论文和承担科研项目及取得成果........................................83

致谢............................................................................................................................85

第一章绪论

1.1 引言

砌体结构是最原始的建筑结构，中国的砌体结构在历史的长廊中始终熠熠生

辉。举世闻名的万里长城，记载了超过两千年的文明史；建于隋朝大业年间的河

北赵县赵州桥，为世界上最早的空腹式石拱桥，也是典型的砌体结构，如图 1-1

所示。

(a) 万里长城 (b) 赵州桥

图1-1 古老的砌体结构

新中国成立后，砌体结构得到了快速的发展，全国约有90%左右的基础工程

建设采用砌体作为墙体材料[1]。民用建筑特别是住宅建筑中广泛采用砌体墙承重，

工业建筑中也普遍用砌体墙承重。由于砌体结构砌筑方法多样性、结构形式多变

等原因，加之结构长期服役期间，不断受周围环境影响和各种荷载的作用，致使

结构或者构件产生了不同程度的损伤。这些因素使得结构的力学参数发生了变化，

对其进行性能分析和评价相当困难。

弹性模量是砌体结构的基本力学性能指标和结构安全性的重要指标，同时也

是砌体构件强度、刚度以及抗震性能评估时必不可少的一个参数。过高或过低的

估算砌体结构弹性模量都会造成计算结果的不准确，因此弹性模量值的精确度对

砌体结构的有限元分析与计算是非常重要的。课题旨在通过基于环境激励的动力

测试方法获取砌体试件的基本频率，结合有限元计算砌体试件的弹性模量，并以

此弹性模量作为有限元的输入参数进行砌体试件的损伤识别。

结构的力学参数可分为静力参数和动力参数，分别通过静力测试方法和动力

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

15 积分 4人已下载

立即下载 VIP免费下载

摘要：

摘要砌体结构是一种历史悠久的建筑结构形式。在长期的使用过程中，环境侵蚀、自身老化、自然灾害因素等联合作用，将导致其材料的损伤累积、构件和结构承载能力下降。通过无损的检测方法获取动力特性，进而识别材料的损伤情况是当前研究的热点。弹性模量是砌体结构安全性的重要指标，也是砌体结构损伤识别研究及有限元分析的基础参数。为通过无损方法获得此重要参数，首先通过对砌体试件进行环境激励与人工激励的动力测试，获取砌体试件的基本频率，并验证环境激励识别砌体试件基本频率的可行性。在此基础上，通过采用马尔科夫蒙特卡洛（MarkovChainMonteCarlo,MCMC）抽样的贝叶斯方法，获得砌体试件基本频率的后验概率...

展开>> 收起<<

基于贝叶斯方法的砌体材料损伤识别.pdf

共87页,预览9页

还剩页未读，继续阅读

基于贝叶斯方法的砌体材料损伤识别

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

推荐作者

热门标签

举报选择: