Damage sensitivity evaluation of vibration parameters under ambient excitation

Abstract

Traditionally, damage detection techniques in bridges have focused on the use of modal-based damage features as they are supported by a well established theoretical base and can be directly related to changes in stiffness. However, modal-based damage features can also be difficult to extract from noisy data and thus their application may require initial large amplitude excitation, in the form of impact tests on closed bridges, to extract an accurate and full modal response. Furthermore, as damage detection and identification methodologies evolve to account for more and more variables, yet still fail to make the breakthrough into industrial applications, there may be an argument to reassess the selection of features used as inputs to these methodologies. Following some recent investigations that discovered a correlation between certain vibration parameters and bridge condition, the present study expands this inquiry by examining the level of damage sensitivity exhibited by various vibration parameters under more scrutiny. Initial sections of this paper outline the vibration parameters to be assessed and the damage detection algorithms employed. The paper finishes with the assessment of the chosen vibration parameters by using vibration data from a progressively damaged bridge subjected to ambient excitation. The paper also suggests a methodology that can be implemented to reduce uncertainty regarding ambient excitation sources.