Detection and quantification of damage in bridges using a hybrid algorithm with spatial filters under environmental and operational variability

It is essential to isolate the environmental effects on the structure from the incipient damage during structural health monitoring, failing which, may mislead the diagnostic process in a way, either, by giving false positive alarms or masking the existing real damage. The modal filter is popularly used to handle environmental variability, while detecting the current state of the structure, during structural health monitoring. However, this technique is a qualitative one and it cannot identify the spatial location and the extent of the damage. In this paper, the modal filter is combined with an inverse algorithm to localize and quantify the extent of damage, while handling environmental variability. The inverse damage detection problem is formulated as a constrained optimization problem and solved using a Multi cluster hybrid adaptive differential search algorithm (MCHADSA). A new damage index is proposed, to detect the exact time instant of damage, alleviating the effecting confounding factors like environmental and operational variability (EoV) and measurement noise. Numerical experiments are conducted to evaluate the performance of the proposed inverse damage diagnostic MCHADS algorithm and the results are presented in this paper. A beam girder is taken as the first example followed by a more realistic example of a live bridge across river Amaravati near Dharapuram, Tamil Nadu, India. The studies presented in this paper indicate that the proposed Modal filter-based hybrid inverse algorithm is effective in localizing as well as quantifying the damage. The effect of modeling errors is also investigated in the proposed algorithm.