Modeling same-direction two-lane traffic for bridge loading

Many highway bridges carry traffic in two same-direction lanes, and modeling the
traffic loading on such bridges has been the subject of numerous studies. Different
assumptions have been used to model multiple-presence loading events, particularly
those featuring one truck in each lane. Using a database of weigh-in-motion
measurements collected at two European sites for over 1 million trucks, this paper
examines the relationships between adjacent vehicles in both lanes in terms of vehicle
weights, speeds and inter-vehicle gaps. It is shown that there are various patterns of
correlation, some of which are significant for bridge loading. A novel approach to the
Monte Carlo simulation of such traffic is presented which is relatively simple to apply.
This is a form of smoothed bootstrap in which kernel functions are used to add
randomness to measured traffic scenarios. It is shown that it gives a better fit to the
measured data than models which assume no correlation. Results are presented from
long-run simulations of traffic using the different models and these show that
correlation may account for an increase of up to 8% in lifetime maximum loading.