基于灰色理论的钢桥疲劳寿命评估

发布时间：2018-09-12 12:41

【摘要】：随着经济的发展,我国建成了为数众多的桥梁,作为控制桥梁使用寿命的重要因素之一,现如今疲劳问题越来越多地被人们所关注。对桥梁进行疲劳寿命研究不仅有助于延长桥梁使用年限,而且能为定期进行维护提供理论支撑。采用传统方法研究桥梁构件的疲劳问题时,需要采集大量的应力历程数据,这使得研究工作效率低下,而应力谱数据不足又会造成评估结果出现很大的偏差。与此同时,灰色理论在预测方面的成效愈发显著,它能在少数据、贫样本的基础上对机械结构和部件的疲劳寿命进行有效预测。从这一角度出发,将灰色理论引入到钢桥的疲劳寿命研究中,对不同构件进行预测。本文借助于Matlab和Midas Civil软件,在现有钢桥疲劳分析手段以及预测评估方法的基础上,对新菜市桥关心构件的疲劳损伤度和剩余疲劳寿命进行了一系列估算。引入灰色理论,以损伤度为基础建立灰色预测模型,采用4种不同的方法进行相关预测,主要内容如下： (1)简要介绍了传统线性累积损伤理论评估钢桥剩余疲劳寿命的具体步骤,并依此为基础确定了新菜市桥疲劳损伤分析的基本思路。 (2)通过Midas软件建立该桥的分析模型,对不同杆件进行应力影响线求解,分析并选取桥梁的关心杆件；加载虚拟车辆荷载,统计4种杆件在不同加载时间后的应力历程；编谱并计算不同杆件不同加载时间下的疲劳损伤度和剩余寿命。结果显示,相同的加载时间下,4种杆件的累积损伤度基本处于同一个数量级,但4号斜杆的损伤度更大,因而疲劳寿命更短。 (3)基于传统方法计算得到的损伤度建立标准GM(1,1)模型进行损伤度评估,结果显示：拟合精度在90%以上,但进行预测时误差偏大。改进模型,尝试降低误差值,通过计算分析可以发现：基于arccot(.)变换的GM(1,1)改进模型,不仅拟合精度有所提高,而且预测值偏差也较小,因而该模型更加有效；而其余两种模型在提高拟合精度和降低预测值误差方面均未得到明显改善。综上所述,在钢桥剩余疲劳寿命和损伤度的评估中,灰色理论能够取得一定的效果,并且模型的改进有助于提高预测的精度。
[Abstract]:With the development of economy, a large number of bridges have been built in China. As one of the important factors to control the service life of bridges, the fatigue problem has been paid more and more attention. The study of fatigue life of bridges can not only prolong the service life of bridges, but also provide theoretical support for periodic maintenance. When the traditional method is used to study the fatigue problem of bridge members, it is necessary to collect a large amount of stress history data, which makes the research work inefficient, and the lack of stress spectrum data will lead to great deviation of the evaluation results. At the same time, the grey theory is more and more effective in predicting the fatigue life of mechanical structures and components on the basis of a small number of poor samples. From this point of view, the grey theory is introduced into the fatigue life of steel bridge, and the different components are predicted. With the aid of Matlab and Midas Civil software, based on the existing fatigue analysis methods and prediction and evaluation methods of steel bridges, a series of estimation of fatigue damage degree and residual fatigue life of concerned members of Xincaishi Bridge are carried out. The grey theory is introduced, and the grey prediction model is established on the basis of damage degree. Four different methods are used to predict the damage degree. The main contents are as follows: (1) the concrete steps of traditional linear cumulative damage theory for evaluating residual fatigue life of steel bridges are briefly introduced. Based on this, the basic idea of fatigue damage analysis of Xincai Bridge is determined. (2) the analysis model of the bridge is established by Midas software, the stress influence line of different members is solved, and the concerned members of the bridge are analyzed and selected. After loading the virtual vehicle load, the stress history of four kinds of bars under different loading time is calculated, and the fatigue damage degree and residual life of the four kinds of bars under different loading time are calculated. The results show that the cumulative damage degree of the four kinds of bars is basically in the same order of magnitude under the same loading time, but the damage degree of the No. 4 inclined bar is more serious. So the fatigue life is shorter. (3) based on the damage degree calculated by the traditional method, the damage degree is evaluated by the standard GM (1 / 1) model. The results show that the fitting accuracy is more than 90%, but the error in prediction is too large. Improve the model, try to reduce the error, through the calculation and analysis can be found: based on arccot (.) The transformed GM (1 / 1) improved model not only improves the precision of fitting, but also reduces the deviation of prediction value, so the model is more effective, while the other two models have not been improved obviously in improving the fitting accuracy and reducing the error of prediction value. To sum up, the grey theory can obtain certain results in the evaluation of residual fatigue life and damage degree of steel bridges, and the improvement of the model is helpful to improve the accuracy of prediction.
【学位授予单位】：大连交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U441.4

【相似文献】