IDENTIFICATION OF BRIDGE FRAGILITY PARAMETERS USING SEISMIC DAMAGE DATA OBTAINED FROM EXPERIMENTAL STUDY

Open Access
Author:
Chi, Chao
Graduate Program:
Civil Engineering
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
November 14, 2011
Committee Members:
  • Swagata Banerjee, Thesis Advisor
Keywords:
  • fragility parameter
  • EKF
Abstract:
The study describes a procedure to identify bridge fragility parameters utilizing its vibration response recorded during experimental study. For this purpose, bridge damage data observed in a near full-scale shake table experiment is utilized. The bridge was tested under a sequence of earthquake ground motions with increasing intensities. Low and high amplitude tests were performed in series to observe the seismic performance of the bridge starting from yielding to complete failure. In the present study, recorded bridge acceleration during high amplitude tests is utilized and further analyzed to evaluate the degraded performance of the bridge after each high amplitude test. This is done by using extended Kalman filtering (EKF) technique as a tool. The degraded performance of the bridge after each run is measured in terms of effective stiffness of the bridge at pier ends. In parallel, finite element (FE) model of the same bridge is developed in order to perform time history analysis under a set of earthquake ground motions with various hazard levels. Before applying the ground motions, the FE model is updated with the effective stiffness of the bridge obtained from EKF after each high amplitude test. This is important to numerically simulate the post-damaged condition of the bridge and to quantify the gradual progression of bridge damage when subjected to earthquake ground motions in sequence. After each time history analysis, bridge response is obtained in terms of the rotation at bridge pier ends. Thus obtained response from time history analyses is used for fragility curve development. The change in fragility parameters represents the progressive damage of the bridge when subjected to ground motions with incremental intensity. Statistical uncertainty of the fragility curves is measured in terms of 90% confidence interval.