Local Nonlinear Dynamic Analysis Based On Reduced Model

Recently several types of nonlinear component or energy-dissipating component are often pre-assembled into building structures to isolate or dissipate earthquake energy. The use of such energy-dissipating components leads to huge nonlinear computation for structural analysis and corresponding nonlinear time-history analysis of large complex structure is very time consuming. Therefore, it is necessary to find a reasonable reduced model instead of the original one to speed the nonlinear computation. In the present thesis, dynamical response of structure with local nonlinearity is studied by using the model reduction method based on local rigid body mode and fast nonlinear analysis (FNA) method. The technique presented here is applied in the structural dynamic response analysis of a high-rise building with local nonlinearity in the end. The work of this dissertation includes the following three parts:(1) Accuracy and efficiency of three methods, such as classical static model reduction method, iterative IRS method and the model reduction method based on local rigid body mode, are compared. The effect of position of boundary nodes of rigid body region on accuracy of reduced model is studied for two cases:a flat plate and a space framework.(2) The fast nonlinear analysis method is proposed for the reduced model. The linear components and nonlinear energy-dissipating components are modeled separately. The reduced model of linear components is obtained by using the model reduction method based on local rigid body mode. The nonlinear forces are introduced to take the place of nonlinear components by using FNA method. Precise time integration method is used to complete the structural dynamical analysis.(3) A program based on the platform of MATLAB and ANSYS is presented to get the reduced model and complete calculation of dynamical response of structure with local nonlinearity. The program is used in the dynamical analysis of a real high-rise building with local nonlinearity. The computation efficiency and scope of application of this program is discussed based on this case.