Modeling of an Instrumented Building Subjected to Different Ground Motions

Abstract

The purpose of this paper is to compare the recorded response of an instrumented building in San Jose, California, with the response obtained from a finite element model of the building. Three ground motions from past earthquakes, one with significantly higher-amplitude than the other two, are used for the study. A detailed model of the building is created and calibrated using one of the lower-amplitude motions. It was demonstrated that the model closely simulates the building response under the other low-amplitude motion, while the results obtained using the larger motion did not match well with the recorded motions. For the higher-amplitude shaking, the building was more flexible as its fundamental period increased because of rocking effects. Hence, the model has to be updated by including soil-structure interaction effects using more flexible translational springs and adding vertical springs; resulting in a good match between measured and computed motions. As part of the study, the variation of the dynamic characteristics of the building obtained from a modal analysis of the recorded motions was investigated too. It can be concluded that a calibrated model based on only low level recorded motions may not be suitable to a higher level ground shaking.