Study On Mechanical Properties Of Beam-to-column Weak-axis Connections With Gusset Plate Based On Component Method And Global Sensitivity

At present,the research on beam-to-column connections of steel frames mainly focuses on strong-axis connections,and the research on weak-axis connections is relatively lacking.The double web bottom angle steel beam-column weak axis connection joint with joint plate is a kind of joint that can realize full assembly connection on site.The nodal plate is welded to the flange of the beam,and the beam with nodal plate is connected to the column by angle steel and high-strength bolts.In engineering,this kind of assembled node is a commonly used form of connection.Therefore,this thesis studies such weak axis connection joints.In the process of parametric analysis of joints using ABAQUS finite element software,it is found that when only a certain parameter is analyzed,multiple models need to be established for comparative study.If there are many parameters,more models need to be established for analysis.If these models are modeled one by one,it will take a lot of time and effort,and it is prone to errors.In this thesis,based on Python language,the automatic modeling and post-processing program of double-web bottom angle steel beam-column connections with nodal plates is developed,and the automatic modeling and post-processing are realized by docking with ABAQUS.Through finite element analysis,it is found that when the model is more complex,the calculation time will be too long,which may exceed the processing capacity of the computer.In view of the above problems,this thesis uses genetic algorithm to optimize BP neural network,so that it has good adaptability and generalization ability.It is used to predict the output response such as the initial rotational stiffness of the joint and improve the calculation efficiency.In this thesis,the parametric analysis of single parameter change of joints is carried out by Python and ABAQUS docking.On this basis,the local sensitivity analysis of the weak axis connection joint of double web bottom angle steel beam column with gusset plate is carried out.The parameters that have great influence on the initial rotational stiffness of the joint are selected.Considering the influence of the correlation between the node parameters,the weights and thresholds of the BP neural network are optimized based on the genetic algorithm.The global sensitivity analysis is carried out on the parameters that have great influence on the mechanical properties of the joints.The influence of simultaneous variation of multi-dimensional parameters on the initial rotational stiffness of weak-axis connection joints of double-web bottomangle steel beams with gusset plates was studied.It is found that the initial rotational stiffness of the joint decreases with the increase of column section height,and increases with the increase of column web thickness,beam section height,column flange thickness,bolt diameter and upper angle steel thickness.Combined with the results of global sensitivity analysis,the thickness of the column web,the height of the column section and the corresponding sensitivity coefficient are introduced into the component method,and the initial rotational stiffness formula of the joint based on the component method is modified by Matlab.Comparing the calculation results of the modified formula with the finite element results,the relative error between the two is small,which verifies the accuracy of the modified formula,and shows that the formula has certain reference value for engineering application.