LOCAL DAMAGE DETECTION OF SIMPLY SUPPORTED BEAM ON THE EXCITATION OF MOVING LOAD BASED ON AVERAGE CURVATURE MODAL

Based on the average curvature modal and gapped smoothing method, a new algorithm to detect the local damage of a simply supported beam under the excitation of moving loads is proposed. Without need for the dynamic response data before a structure was damaged, this algorithm has good noise immunity, and could be used in the situation of multiple moving loads. Firstly, by calculating the dynamic response of a simply supported beam under the excitation of a mass-spring-dashpot vehicle system using Newmark-βtime domain integral algorithm, the ascendancy of a basic mode in the response is confirmed, and this explains the meaning to average the curve modal. Thereafter, an FEM model including a moving mass-spring-dashpot vehicle system on a simply supported beam is built to conduct a numerical simulation. Here a triangle fitting function is constructed to fit the average curvature modal. The numerical simulation shows that the algorithm is effective in the situation of multiply damage location, multiply moving loads, and noise environment. This algorithm provides a new research thought for the damage detection of bridges opened in normal traffic.