低周荷载下型钢混凝土柱抗剪性能智能分析

Intelligent analysis of shear performance of shape-rolled concrete column under low-cycle loading

为了研究型钢混凝土短柱的抗震性能,通过21根SRC柱在低周反复荷载作用下的剪切黏结破坏试验结果,在利用灰色关联理论进行模型输入变量选择的基础上,建立了其基于改进神经网络模型的受剪承载力分析模型,进而研究了混凝土抗压强度、腹板宽度、轴压比及箍筋间距等因素对SRC柱抗剪能力的影响规律,并对SRC柱的受剪承载力进行了预测。结果表明,所提方法能够反映受剪承载力与影响因素间的非线性变化规律,可供SRC柱在低周往复荷载作用下的受剪承载力分析参考。

In order to study the shear capability of shape-rolled concrete （SRC） column with bonding failure under cyclic loads, a method, based on improved neural network model with the Levenberg-Marquardt optimized algorithm, is proposed to analyze the shear capability. In this method, gray relational theory is introduced for analysis of factors affecting the shear capability, such as material strength, geometry dimension, shear-span ratio and axialcompression ratio. This method is checked up by means of test data set. The obtained results show that numerical data are in good agreement with the experimental results. So it could be used to analyze the shear capability of SRC column with bonding failure under cyclic loads.