Model Based Pavement-Vehicle Interaction Simulation for Life Cycle Assessment of Pavements

Responsible for about a third of the annual energy consumption and greenhouse gas (GHG) emissions, the U.S. transportation Network needs to attain a higher level of sustainability. This is particularly true for the roadway Network and the design of pavements in it. Improving sustainability of this network necessitates a fundamental understanding of pavements and their interaction with the users, vehicles. Vehicle fuel consumption required to overcome resisting forces due to pavement-vehicle interaction (PVI) is an essential part of life-cycle assessment (LCA) of pavement systems. These PVIs are intimately related to pavement structure and material properties. While various experimental investigations have revealed potential fuel consumption differences between flexible and rigid pavements, there is high uncertainty and high variability in the evaluated impact of pavement deflection on vehicle fuel consumption. This research adopts the perspective that a mechanistic model which links pavement structural and material properties to fuel consumption can contribute to closing the knowledge gap of PVI in pavement LCA. With this goal in mind a first-order mechanistic pavement model is considered; a Bernoulli-Euler beam on viscoelastic foundation subjected to a moving load. Based on the model, scaling relationships are developed between the input parameters of top layer and subgrade moduli, pavement thickness, and loading conditions, with their impact on PVI and vehicle fuel consumption. The strength of these scaling relationships and their ability to guide pavement design are presented through examples. An original calibration-validation method is established through wave propagation, using Falling Weight Deflectometer (FWD) time history data from FHWA's Long Term Pavement Performance program (LTPP). Statistical analyses of model parameters are performed on pavement material properties (top layer and subgrade moduli), structural properties (thickness), and loading conditions; obtained from model calibration and the LTPP datasets for 5643 FWD points, representing the U.S. roadway Network. These distributions are used as inputs to Monte Carlo simulations to determine the impact of flexible and rigid pavement deflection on passenger car and truck fuel consumption within the roadway Network. From the Monte-Carlo simulations it is shown that rigid pavements within the Network behave better than flexible ones in regard to PVI due to higher stiffness. A comparison of the deflection induced PVI on flexible and rigid pavements with independent field data provides a reality check of the order of magnitude estimates of fuel consumption, as determined by the model. Moreover, distributions of model based change in fuel consumption are used in a comparative partial LCA of flexible and rigid pavements. It is shown that the impact of deflection induced PVI (at 95% confidence intervals) becomes increasingly important for high volume flexible roadways and can surpass GHG emissions due to construction and maintenance of the roadway system in its lifetime.