Bayesian operational modal analysis and Markov chain Monte Carlo-based model updating of a factory building

Abstract

This paper presents the results of a full-scale ambient vibration test, modal analysis and model updating of a typical 14-story reinforced concrete factory building in Hong Kong. A 12-setup test was conducted in the building’s three staircases using six tri-axial accelerometers. The modal parameters of each setup were identified following the Bayesian approach and the partial mode shapes from different setups were assembled using the least-squares method. The factory building was then modeled as a shear building and the Markov chain Monte Carlo (MCMC)-based Bayesian model updating method was applied utilizing the identified modal parameters to determine the probability density functions of the various inter-story stiffness values. Four classes of shear building models were studied and the MCMC-based Bayesian model class selection was developed to identify the most plausible model class. The identified modal parameters and model updating results were analyzed and are discussed in detail. This study provides valuable experience and information for the future development of the structural model updating and structural health monitoring of building systems.