Predicting Roughness Progression Of Asphalt Pavements By Empirical-Mechanistic Model

The evaluation of road roughness progression represents one of the most important issue of modern pavement design and management. The roughness deterioration models developed so far, are often based on an empirical approach. Usually, an extensive long-term measurement campaign on a wide range of pavement sections is needed to calibrate such complex regression models and additional difficulties occur in collecting data on overlay roughness evolution. Furthermore, empirical models are unable to take into account new pavement materials performance. Empirical-mechanistic approach could partly solve these problems allowing the development of deterioration models by numerical simulations. In this paper an empirical-mechanistic model, implemented in a computer software, is described in order to simulate the progression of pavement roughness as function of loads, climate, structural configuration, overall construction quality and material properties. The analysis is performed by dividing the road section into 0.25÷0.33 m long sub-sections. Material properties and layer thickness are assigned to the sub-section through a autoregressive process and non linear elastic behaviour of granular materials is taken into account. The simulation is performed through an iterative process according to which, within each computational step, the program calculates pavement stress-strain level and permanent deformation induced by dynamic loads, for 16 different types of axes and 8 daily climatic condition in each sub-section. At the end of step the total rutting is computed in each sub-section, the profile is updated and the IRI index of the road section is evaluated. The model has been used to simulate roughness progression and a preliminary parametric analysis has been performed whose results are briefly illustrated.