| Full Text | Home |
A Data-Driven Approach for Earthquake Damage Detection Burcu Gunes
Assistant Professor, Department of Civil Engineering, Atılım University, Kızılcaşar Koyu, İncek, Ankara, TURKEY, burcu_gunes@atilim.edu.tr
Dionisio Bernal
Associate Professor, Department of Civil and Environmental Engineering, Northeastern University, Boston, MA 02115, USA, bernal@neu.edu
Abstract
The need for a rapid assessment of the state of the critical and conventional civil structures have been demonstrated during recent earthquake disasters. Considerable research efforts have been directed towards the use of vibration measurements and numerous methods have been proposed to assess structural health by comparing the vibration signals measured on the structure before and after damage, with both signals analysed using linear methods. This paper, however offers an alternative means of assessing the damage by estimating the degree of non-linearity present within the given response measurement. The error between the recorded data and the predicted response under the premise of linear behavior calculated using eigensystem realization algorithm with observer Kalman filter is used as a criteria to assess the state of the structure. The amount of the input energy dissipated through hysteretic action is used as the response parameter for determining the level of damage. The numerical simulations are carried out with realistic levels of noise on a 4-story building with members having a Bouc- Wen type restoring force-displacement relationship and the method is found promising as an initial screening process for assessing the health of structural systems after an
extreme event.
|