Damage Detection for Framed RCC Buildings using ANN Modeling

Abstract

A structure is evaluated, after an earthquake, to find out whether it is usable or requires repair and retrofitting. The natural time period, damping, and mode shape are the primary dynamic characteristics of any structure that is related to seismic forces during an earthquake event. Periodic monitoring using vibration measurements is one of the most effective nondestructive methods to identify the damage level. A significant deviation in natural frequencies or damping ratios from the undamaged states indicates the possible occurrence of damage. In the present study, the dynamic characteristics of a 2D rigid frame of a reinforced concrete building are obtained analytically under different levels of damage. Further, using an artificial neural network (ANN) model the authors have developed a correlation between the damage in the frame of the reinforced concrete building with its known dynamic characteristics. Use of such correlations for evaluating the health status of the buildings has been discussed.