摘要
针对实际工程中钢结构锈蚀数据采集周期长、数据少且离散性较大等问题,通过对Q235钢材进行模拟酸雨喷淋加速锈蚀试验,得到钢材的锈蚀质量损失数据。结合随机过程理论,建立了基于时变Gamma随机过程的钢材锈蚀质量增长模型。然后利用Bayesian更新理论和马尔卡夫链蒙特卡洛方法,对钢材锈蚀质量增长模型随机参数进行更新,进而预测钢材在下一个试验周期内的锈蚀质量增长轨迹。结果表明,该方法可以实现钢材锈蚀质量增长随机过程模型随机参数的修正,有效预测钢材锈蚀质量增长规律。可为预测钢材锈蚀寿命和制定钢结构维护策略提供参考。
Aiming at the problems of long collection period,small amount and large dispersion of steel structure corrosion data in actual engineering.The accelerated corrosion test by simulating acid rain spray on steel Q235 was carried out to obtain the corrosion mass data.Combined with the stochastic process theory,a steel corrosion mass growth model based on the time-varying Gamma stochastic process was established.Using Bayesian update theory and Markov chain Monte Carlo method,the stochastic parameters of the steel corrosion mass growth model were updated,and the growth trajectory of the steel corrosion mass was predicted.The results show that this method can modify the random parameters of the stochastic process model of steel corrosion quality growth and effectively predict the law of steel corrosion quality growth.It has important reference value for predicting the corrosion life of steel and formulating maintenance strategies for steel structures.
作者
方苇
陈梦成
张锐
谢力
FANG Wei;CHEN Mengcheng;ZHANG Rui;XIE Li(The State Local Joint Engineering Research Center for Security Technology of Operation Maintenance in Rail Transit Infrastructures,Nanchang 330013,China;School of Civil Engineering and Architecture,East China Jiaotong University,Nanchang 330013,China)
出处
《建筑结构学报》
EI
CAS
CSCD
北大核心
2020年第S02期382-388,共7页
Journal of Building Structures
基金
国家自然科学基金项目(51878275)
江西省教育厅科学技术研究项目(GJJ190299)
