Kalman filter based estimation of neutral axis position of bridge deck under traffic loading

Abstract

In recognizing that small structural defects could be often discerned from the variation in strain or its derivatives rather than in acceleration, strain-based methods for structural damage detection have received increasing attention recently. In this study, the neutral-axis position of bridge deck cross-sections is proposed as a damage indicator for bridge deck assessment and a Kalman filter (KF) based method for optimal estimation of the neutral-axis position from dynamic strain measurement data under operational traffic is developed. As observed from the monitoring data, under traffic effect, bridge deck performs like a flexural beam, i.e., the deck top compresses and the deck bottom extends concurrently, or vice versa. Based on the relationship between the neutral-axis position and strain response, a KF estimator for locating the neutral-axis position is formulated and used to verify its robustness to noise disturbance through numerical simulations. The numerical studies show that the estimator generates satisfactory results in the presence of noise. The proposed K.F estimator is further applied for neutral-axis estimation of the suspension Tsing Ma Bridge (TMB) using long-term monitoring data of dynamic strain responses. The results show that the neutral-axis position is insensitive to the traffic environment and hence can serve as an indicator for deck condition assessment.