Experimental validation of a modal flexibility-based damage detection method for a cyber-physical system

Abstract

The detection and localization of damage in a timely manner is critical in order to avoid the failure of structures. When a structure is subjected to an unscheduled impulsive force, the resulting damage can lead to failure in a very short period of time. As such, a monitoring strategy that can adapt to variability in the environment and that anticipates changes in physical processes has the potential of detecting, locating and mitigating damage. These requirements can be met by a Structural Control (SC) system that is capable of measuring and analyzing dynamic responses in real time. Such a system requires an effective damage identification method capable of locating and quantifying damage. The Eigenparameter Decomposition (ED) of Structural Flexibility Change method is validated with real data and can be used in the computational core of this system. The condition screening is implemented on a damaged structure and compared to an original baseline calculation, hence providing a supervised learning environment. A laboratory-scale experimental study on a 5-story shear building with different damage conditions subjected to an impulsive force has been chosen to validate the effectiveness of the proposed method. The effectiveness of the ED method to locate damage is compared to that of the Damage Index method. The ED method has also been modified to be able to locate and quantify damage. The experimental study shows the effectiveness of the modified method.