Statistical structural health monitoring: Methodologies and Applications

Abstract

Innovative methodologies of statistical structural health monitoring (SHM) and their applications are presented in this thesis. First, Hotelling's T2 control chart is used to construct a multivariate extension of a time series based SHM scheme. The sensitivity of the defined damage indicator is further improved by invoking multivariate exponentially weighted moving average (MEWMA) control chart analysis. Based on this MEWMA control chart analysis and the vibration responses representation by sample principal component coefficients (PCCs), structural damage detection is then addressed via a multivariate statistical approach. Finally, a nonparametric statistical SHM framework is formulated. The efficacy of the methodologies is demonstrated by a series of applications including numerical examples of a progressively damaged reinforced concrete (RC) frame, a five-story shear frame, a shear wall, a twenty-degree-of-freedom system and a hyperbolic paraboloid roof shell as well as the experimental example of the I-40 Bridge benchmark.