Abstract:Differences in vehicle load and its response on multi lane highway bridges were crucial to bridge design and related researches. These differences, especially the load effects with long loading length, were poorly illustrated by the current multi lane traffic load models. The 28 day free flow weigh in motion(WIM) data on a two lane highway and a four lane highway were offered. The data were obtained from the year 2009 to 2013. Based on the data, characteristics in the vehicle lane selection and spatial load distribution were studied. Multi lane load response with long loading length was analyzed. A new multi lane load model was proposed, based on which model coefficients values were calibrated. The results suggest that statistical stability of the vehicle and load analysis can only be ensured with a data length over 3 weeks. Vehicle and load distribution in each lane was of significant difference, indicating that the equal distribution assumption of vehicle lane load in current Chinese code was in need of modification. Variation of the vehicle load was substantial while the maximum load values were of not much difference, which suggests that the current multi lane load model can no longer satisfy the modification requirements raised by varied loading lengths. The load response ratios of different lanes with long loading length roughly remained as constants, indicating that extreme load responses in different lanes were different and would not appear simultaneously. The recommended conceptual multi lane load model provided a reasonable and scientific elaboration on the relationship of extreme load responses among different lanes and all lanes with long loading length. The model proved to be reasonable and of more engineering application value by the coefficient calibration.