Detection and estimation of damage in framed structures using experimental modal data

Abstract

The inevitable ageing and degradation of buildings and the structural failures that follow, have ignited a need for early prognosis of probable structural failures so that proactive measures can be undertaken. Hence, one of the important steps of structural health monitoring (SHM) process is the detection of damage and estimation of damage severity. Modal data can be effectively used for this purpose owing to their sole dependency on mechanical characteristics of a structure. However, the focus of mode shape-based damage detection techniques has concentrated only on symmetric structures whereas the existing buildings are typically asymmetric. This study presents a damage detection methodology using the behaviour of mode shape derivatives such as mode shape slope and mode shape curvature for a symmetric framed structure applied on an experimental model tested using a shaking table, and a calibrated finite element model. Furthermore, an extended parametric analysis has been performed to investigate damage localization and quantify severity. Finally, the models have been modified to incorporate the irregularity effects and damage detection possibility has been explored. The study enables to provide key conclusions for damage detection with respect to localization and severity in the steel frame model. Damage detection method using the mode shape curvature is identified to be more sensitive as opposed to mode shape slope method. And the effect of mass irregularity on the detection methods were identified.