Moving Force Identification as a Bridge Damage Indicator

Information on the condition of bridges is primarily obtained through the use of visual inspection methods. These methods are unreliable due to an overdependence on human judgement and also inconsistencies due to human objectivity. A scientific approach is a more suitable alternative. Some authors use changes in the natural frequencies of the bridge to indicate possible damage but these methods aren't suitable for local damage detection. Mode shapes have also been used but are more difficult to infer from measurements. This paper investigates the use of a Moving Force Identification (MFI) algorithm in conjunction with bridge deflection data. MFI back-calculates a vehicles applied axle forces to a bridge. It has been found that damage in a bridge changes the calculated axle forces substantially. These calculated axle force histories are used to infer damage from. The damage indicator used here is based on a linear regression analysis of the axle force histories. It is found that the absolute value of the slope of the linear regression fit increases with damage. Hence, by monitoring this parameter, information on possible bridge damage may be supplied on a vehicle by vehicle basis.