Traffic load patterning on long span bridges: A rational approach

Abstract

For highway bridges, current live load models from design standards typically replicate the effects of real traffic streams at a certain return period. Most standards stipulate that the worst load effect should be considered, obtained by patterning the live load on the adverse portion of the influence line only. Certainly, this is a worst-case scenario, but the probability of this occurring is surely very low and daily traffic does not pattern in this manner. This paper examines design standards’ approach to load patterning and proposes a relatively simple approach to better reflect the reality of traffic load patterning. As an illustrative example, traffic microsimulation is adopted to simulate the traffic flows on a hypothetical 1200 m bridge. We consider the possible road topologies that might give rise to design standard loading patterns, showing that it is not possible to achieve the code-implied traffic pattern in reality. The load effects results are extrapolated to 5, 75 and 1000 year return periods, representing typical assessment and design return periods respectively. From these results, a rational and practical load patterning method is proposed, based on traffic topologies obtained from an adverse combination of congested and free flow traffic. It is applied to two case study bridges and the results indicate that our proposal can reduce the traffic loading demand by up to 45%. Such reductions can be invaluable in the design of super-long span bridges, or the assessment of existing bridges.