| At present,the confinement effect between steel tube and concrete is different,so there are some limitations.However,the neural network method can find out the nonlinear relationship between input and output by learning and training the sample data without any assumptions,and predict the new input variables accurately.In addition,the efficient and scientific method of neural network is suitable for the computer simulation technology,which can realize the visualization of the entire process of the deformation and damage of the components,which can help to solve the practical engineering problems.The failure modes of concrete-filled rectangular steel tubular columns are mainly divided into two types: strength failure and instability,and furthermore,the strength failure is divided into local buckling failure and integral bending failure according to the final failure manifestation.It is concluded that the smaller the aspect ratio(length /width)is,and the bigger the concrete strength is,the bigger the slenderness ratio of the column will be.The bigger the steel strength is,the larger the coefficient of confinement effect,the larger the ratio of the slenderness ratio of the columns which will be damaged by the whole bending.Two kinds of neural network models with different input parameters are established.And then the axial compression stability coefficient is obtained by using the trained neural network model to calculate the axial compressive bearing capacity of different slenderness ratios.The formula of GB50936-2014 is also used to calculate the stability coefficient.Stability coefficient of axial compression in the same slenderness ratio.Finally,the slenderness-stability coefficient curves obtained by the two methods are compared and the conclusion is drawn that the neural network method is more safety than the specification in calculating the axial stability coefficient.The four main theoretical and normative formulas for calculating the bearing capacity of concrete-filled rectangular steel tube columns are introduced.The bearing capacity of axial compression is forecasted,and the axial compression bearingcapacity of corresponding concrete grade is calculated by the formula of GJB4142-2000,GB50936-2014,DBJ29-57-2003.It is concluded that the calculation results of GJB4142-2000 are unsafe in the range of C50 ~ C80;GB50936-2014,the load-carrying capacity-concrete strength curve obtained from DBJ29-57-2003 is in good agreement with the load-concrete strength curve obtained by the neural network method,that is,the superposition theory is adopted to calculate the compressive strength of the high strength concrete filled rectangular steel tube column.Is suitable for the calculation of the axial compressive strength of high strength concrete columns with rectangular steel tube. |