Abstract

Monitoring data obtained at a seismically isolated building in Tokyo during the 2011 off the Pacific coast of Tohoku Earthquake are analyzed to investigate the changes in the modal parameters of the building, which are generally used as global soundness indices in structural health monitoring, in correspondence with the response amplitude of the building. The modal parameters are identified using AR models from each short segment of the record every tien seconds. The AR model orders are selected appropriately by the Bayesian framework. The results show that the natural frequency decreases as the response increases and then regains its value as the response fades, where the value at the end is lower than the initial value. The modal identification is also conducted for daily monitoring data, which are two minute ambient vibration response time histories recorded twice a day, showing that the reduction of the natural frequency is not temporary. This indicates that it is feasible to predict the amplitude of the building response during a severe earthquake from ambient vibration observation before and after the earthquake.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call