Seismic Resilience Assessment of Instrumented Buildings: A Performance-based Monitoring Approach

Abstract

This paper presents a novel probabilistic performance-based monitoring approach for high-resolution seismic damage assessment and resilience quantification in instrumented buildings. The objective is to estimate seismic loss and functionality metrics, which can be integrated into multi-disciplinary community resilience models consisting of interdependent physical, social, and economic metrics. The proposed methodology begins with optimal sensor placement to measure the seismic response of an instrumented building. Then, a structural model-sensor data fusion is implemented using a nonlinear model-based observer to obtain reconstructed engineering demand parameters (REDP) and their estimation uncertainty. The mean and dispersion of REDPs are employed to quantify probabilistic seismic loss consistent with FEMA P-58 methodology, estimating downtime and total loss, which are two necessary parameters for creating component-based functionality curves. These curves define resilience as a function of loss due to extreme events and provide the necessary metrics for linking engineering models with social and economic impact models to estimate multi-disciplinary resilience metrics. The proposed methodology is demonstrated using data obtained from full-scale seismic testing of a wood-frame building conducted at the E-Defense facility in Japan.