Monitoring story stiffness variations in RC buildings under varying seismic intensities using enhanced Hilbert-Huang transform

Abstract

Reinforced concrete (RC) structures are continuously exposed to aging, environmental effects, and extreme events such as earthquakes, resulting in progressive stiffness degradation and increased vulnerability. Accurate, time-resolved assessment of stiffness variations is essential for structural health monitoring (SHM) and seismic performance evaluation. This study proposes a novel framework based on the Enhanced Hilbert-Huang Transform (EHHT), integrating Empirical Mode Decomposition (EMD) and a newly defined Instantaneous Total Amplitude (ITA), to identify story-level stiffness variations. Unlike traditional frequency-domain methods that rely on stationarity assumptions, the proposed approach enables extraction of noise-resilient and physically interpretable stiffness patterns under both linear and nonlinear responses. The dynamic equilibrium equation is reformulated in the time-frequency domain, allowing for robust estimation of stiffness while minimizing the impact of modeling uncertainties, high-frequency noise, and permanent deformations. The method is validated through numerical and experimental studies, including a four-story RC frame with nonlinear behavior and a full-scale five-story RC structure tested on the UCSD-NEES shake table. Comparative analysis with analytical formulations, Power Spectral Density (PSD-based) operational modal analysis, and modal flexibility confirms the superior performance of the EHHT-based method. Findings highlight that stiffness degradation may occur even under weak ground motions, and that characteristics derived from strong shaking may not represent post-seismic conditions accurately. Instead, ambient vibration data recorded after seismic events are more suitable for reliable model updating. The proposed EHHT framework offers a theoretically sound and practically applicable tool for post-earthquake stiffness monitoring in civil infrastructure.