Probabilistic Comparative Analysis of Vehicle-Bridge Interaction Models for Predicting Bridge Response under Moving Vehicles

This paper compares three approaches for predicting the bridge response under moving vehicles. The first approach represents the vehicle as a concentrated moving load, while the second and third approaches simulate the vehicle using single (1-dof) and two (2-dof) degrees of freedom models, respectively. The vehicle-bridge models have been implemented following a finite-difference formulation. These models have been validated against the experimental displacement response of a mid-span prestressed reinforced concrete bridge. Parametric analyses have been conducted to investigate the impact of road roughness on the discrepancies between the models' predictions. An increasing level of road roughness, modeled according to ISO-8608, tends to conceal the mutual differences between the models' outcomes. Furthermore, a Monte Carlo simulation has been employed to fit the probability distributions of two indicators, comparing the bridge displacement responses in three roughness scenarios and considering variable velocity and vehicle characteristics. The primary objective is to assess whether there are scenarios where vehicle-bridge interaction modeling becomes indispensable for accurately estimating the vehicle's effect on the bridge response. The paper presents expressions for four probability density functions of the chosen metric, facilitating the comparison between the 1-dof and 2-dof models with the concentrated load approach.