PRACTICAL ASPECTS IN MOVING LOAD IDENTIFICATION

Several methods have been developed in recent years to identify moving loads on top of a continuous beam using measured vibration responses. The methods can identify the forces with some accuracy, but they have not been tested under field measurement conditions with a bridge–vehicle system. This paper discusses the weaknesses and merits of two methods when applied to a single-span bridge deck. The influence, on the moving load identification, of practical aspects such as measurement noise, sampling frequency, a small number of measured response modes, a small number of measuring points, road surface roughness and non-uniform velocity or braking of vehicle is studied in simulations and experiment. Results show that finite element approach with orthogonal function approximation of the responses give more accurate results, in general, than the exact solution approach for all the studies presented in this paper. The road surface roughness and a large variation in the speed are identified as the two main obstacles leading to erroneous results.