基于实测数据的扁平钢箱温度梯度疲劳应力谱分析方法

A calculation method of thermal fatigue stress spectrum for flat steel girder box based on measurement data

为研究扁平钢箱梁温度疲劳应力谱,以南溪长江大桥悬索桥主梁为研究对象,基于温度传感器长期实测数据,筛选实测数据的日温度极值,运用广义极值模型描述季节极值概率分布并采用极值外推方法得到设计基准期极值模型。引入拉丁超立方抽样(LHS)法对极值模型进行抽样,得到日温度极值样本。结合日温度极值样本和正弦函数模型,构建服役期内关注点的温度时程曲线。基于有限元ANSYS软件平台,分析不同温度梯度下关注点的应力效应,回归温度梯度与疲劳应力的线性关系式,依据温度梯度时程曲线与线性关系式模型,采用雨流计数法得到钢箱梁温度梯度疲劳应力谱。研究表明,模拟抽样生成的温度样本数据符合温度场的季节变化特征,样本概率模型与实测数据概率模型相对吻合。关注点温度梯度疲劳应力谱能够为扁平钢箱梁疲劳寿命设计提供参考。

In order to analyze the thermal fatigue stress spectrum of a flat steel box girder,Nan-Xi Yangtze River bridge was taken as the research object.The temperature data of a steel box girder were collected from a structural health monitoring system(SHMS).Then,the daily temperature maximum values and minimum values were obtained.The seasonal temperature extreme value probability distribution was described by generalized extreme value(GEV)model and the GEV model for a design reference period was established.The method of Latin hypercube sampling(LHS)was introduced in this paper to sample temperature data.Hence,temperature time-history curve in the design reference period by combining temperature extreme sampling and sine function model was obtained.The equation of linear regression between temperature gradient and stress effect was obtained by a finite element analysis.The temperature fatigue stress spectrum of an observation point on the steel box girder can also be obtained based on temperature time-history curve and linear regression model.Research shows that the temperature samples produced by simulation agree with the feature of seasonal temperature change and the correlation between simulation sampling probability model and measured data probability model is good.The thermal fatigue stress spectrum can be used in fatigue life design of steel box girders.