Deterioration assessment of buildings using an improved hybrid model updating approach and long-term health monitoring data

In recent years, it has become increasingly important to develop methodologies for reliable deterioration assessment of civil structures over their life cycle to facilitate maintenance and/or rehabilitation planning processes. Several approaches have been established to address this issue mainly using Bayesian probabilistic model updating techniques with some capability to incorporate uncertainties in the updating process. However, Bayesian model updating techniques are often found to be complex and computationally inefficient as opposed to their deterministic counterparts such as conventional or hybrid techniques of sensitivity-based model updating. Nevertheless, the deterministic model updating techniques have not been well developed for sophisticated assessment applications such as deterioration evaluation. To address these issues, this article presents a novel methodology for deterioration assessment of building structures under serviceability loading conditions, based upon an improved hybrid model updating approach incorporating the use of long-term monitoring data. This is first realized by a simple but effective scheme to simulate the deterioration mechanism in serviceability loading conditions before enhanced with innovative solutions to classify structural elements as well as to handle measurement and updating uncertainties in a meaningful way. The effectiveness of the established methodology is illustrated through a benchmark 10-story reinforced concrete building which is equipped with a long-term structural health monitoring system.