ABSTRACT
Vehicle load is one of most important live loads for long-span bridges to bear. In the case of orthotropic steel box girder, vehicle load will play a very significant role in the long-term fatigue properties of the girder. With the numerous vehicles being loaded repeatedly, the fatigue of the steel box girder becomes an increasingly prominent issue. In order to predict the fatigue properties of orthotropic steel box girder, the real characteristics of vehicle load need to be comprehensively investigated. In this paper, a theoretical research on the technique of dynamic image recognition is performed. With this technique, the transverse distribution of the vehicle load on the bridge deck can be statistically obtained from the video recorded by the bridge monitoring system. Some details of this technique are discussed in the first part of this paper. In the second part, this technique is applied in Sutong Bridge to get the overall traffic flow, as well as its transverse distribution on the bridge deck. Only with the real vehicle load identified can the further mechanical analysis and corresponding maintenance strategy be more accurately and objectively conducted and decided, respectively.
