This article discusses model reduction formulations for the computation of large structural models. The simplest method of model reduction is by using static condensation methods. However, this method might not capture the dynamic properties of the structures. A reduction model based on dynamic analysis is performed to reduce the size of structural computation. Assuming that the damping matrix is in proportion to the mass and stiffness matrix, the free vibration analysis is used as a starting point for the structural model analysis. The transformed matrices are obtained by partitioning the matrices in the equations of motion, considering the retained and condensed degrees of freedom. The retained degrees of freedom can be considered as master degrees of freedom, where the size of the system matrices is expected. By several manipulations, the reduced order model is achieved. The computation starts using Guyan's reduction method, and then the system matrices are updated iteratively. The convergence is defined by comparing the eigenvalues of the successive computations. Numerical examples of four and ten-story shear building models are conducted to show the applicability of the methods. The numerical results show that the reduced-order model obtained using this method can predict the actual model's behavior.

Reduced-order model, dynamic condensation, static condensation, Guyan reduction method, large-scale model