Keywords: long-span, bridge, congested, traffic, jam, traffic flow, weigh-in-motion, WIM, micro-simulation, load model.

It is well established that the critical traffic loading event on long-span bridges occurs in conditions of congested traffic. With theoretical models of congested traffic events, it is difficult to statistically model combinations of heavily loaded trucks and the mix of such trucks with cars and more lightly loaded trucks. Further, predictions of the effect of traffic growth on such congested events are very difficult with existing techniques. This paper presents an alternative approach to the modelling of congested traffic loading events. Micro-simulation, the process of modelling individual vehicles [1], is presented here as a means of predicting imposed traffic loading on long-span bridges.

Simulation modelling is used here to simulate the traffic flow on a congested bridge with a random mix of trucks and cars calibrated to match driver behaviour. Vehicle weights are generated for all vehicles from measured weight histograms obtained using Weigh-in-Motion (WIM) systems. For the assessment of existing bridges, this approach allows the generation of site-specific traffic load models based on realistic traffic characteristics. Simulation reproduces actual traffic flow conditions and the effects of vehicle interactions, considering parameters such as the mix and composition of vehicles in traffic streams. Once calibrated for driver behaviour at the site, the technique permits probabilistic analysis and predictions for accurate bridge load modelling.

The method was tested using a multi-span bridge in the Netherlands. The Hagestein Bridge has 2 lanes of traffic in each direction and an additional lane southbound for cars during the peak hours. It is heavily congested at certain times of the day. The simulation model was developed using knowledge of the traffic volume and the road geometry at the bridge. Vehicle weight information, obtained from a WIM system, was overlain on the vehicle location data generated through simulation. The combined model can simulate realistic vehicle location and weight information for congested conditions. This was used to calculate bending moments and shear forces in the steel plate girder bridge (the middle part of the Hagestein Bridge). Peaks in the simulated strain signal were fitted to a statistical Type I Extreme Value (Gumbel) Distribution and used to determine its statistical properties. These in turn were used to generate strains corresponding to a given return period, i.e., the characteristic values.

From the output obtained it is concluded that micro-simulation software can be used to get reliable site-specific vehicle data generated on the basis of a limited amount of field data. Vehicle classification and weight data, combined with observed driver behaviour, can be used to accurately predict the extent, weight and distribution of vehicle platoons and hence to predict traffic load on long-span bridges.

1

Druitt, S., "An Introduction to Micro-Simulation", Traffic Engineering and Control, London, 649-660, 1999

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £125 +P&P)