高强方钢管高强混凝土长柱偏压性能  

Eccentric Compression Behavior of High-Strength ConcreteFilled High-Strength Square Steel Tubular Long Columns

作　　者：李帼昌[1] 陈博文[1] 杨志坚[1] 邱增美 李晓 刘润泽 LI Guochang;CHEN Bowen;YANG Zhijian;QIU Zengmei;LI Xiao;LIU Runze(School of Civil Engineering,Shenyang Jianzhu University,Shenyang 110168,China)

机构地区：[1]沈阳建筑大学土木工程学院,沈阳110168

出　　处：《建筑钢结构进展》2022年第6期40-53,108,共15页Progress in Steel Building Structures

基　　金：国家自然科学基金(51938009)。

摘　　要：为研究高强方钢管高强混凝土偏压长柱力学性能,基于已有的偏压试验研究,进一步分析了偏压柱破坏形态、荷载-挠度曲线、弯矩-曲率曲线和应变发展规律,探讨了偏心率与长细比对构件塑性发展等力学性能的影响。结合ABAQUS数值模拟研究了不同偏心率的偏压柱工作机理,并由此进一步分析了材料性能参数与构件几何参数对偏压柱的受力性能影响。基于试验与数值模拟,采用《钢管混凝土结构技术规范》(GB 50936—2014)和人工神经网络模型对68个试件的极限荷载P_(u)及其弯矩M_(p)值进行预测。结果表明,《钢管混凝土结构技术规范》(GB 50936—2014)计算的P_(u)值平均比试验与模拟值高2.3%,而计算的M_(p)值则偏于保守且平均偏低13.8%。神经网络模型对P_(u)及M_(p)值的预测结果均与试验和模拟结果吻合较好,验证了神经网络模型的有效性;且该模型可准确预测未知试验结果条件下的试件P_(u)及M_(p)值。To study the mechanical performance of eccentrically compressed square concrete-filled steel tubular long columns incorporating high-strength steel and concrete,the failure mode,load-deformation curve,bending moment-curvature curve,and strain development trend of the columns are further analyzed based on the available eccentric compression tests,and the effects of the eccentricity ratio and slenderness ratio on mechanical behaviors such as the plastic development of the columns are also investigated.The working mechanism of eccentrically loaded columns with varied eccentricity ratios are explored using ABAQUS numerical modelling.The effects of material performance factors and member geometry parameters on column eccentric compression behaviors are further analyzed.Technical Code for Concrete Filled Steel Tubular Structures(GB 50936—2014)and the artificial neural network model are used to predict the peak loads(P_(u))and bending moments(M_(P))of the 68 columns based on tests and simulations.The results indicate that the P_(u)value derived from Technical Code for Concrete Filled Steel Tubular Structures(GB50936—2014)is on average 2.3%higher than the actual and simulated values,while the computed M_(P)value is conservative and 13.8%lower.The neural network model’s predictions of P_(u)and M_(P)values are in good agreement with experimental and simulated data,demonstrating the neural network model’s usefulness.Furthermore,when the test results are unknown,this model can reliably predict the P_(u)and M_(P)values of columns.

关 键 词：高强方钢管 高强混凝土 偏压性能 神经网络 工作机理 

分 类 号：TU398[建筑科学—结构工程]