Download or read book Micromechanical and Probabilistic Approaches to Improve Asphalt Concrete Rutting Performance Prediction Procedures written by Erdem Coleri and published by . This book was released on 2011 with total page pages. Available in PDF, EPUB and Kindle. Book excerpt: Rutting is the load induced permanent deformation of pavements with asphalt concrete (AC) layers. Rutting can occur in the unbound layers of the pavement and in the AC surface layers, the latter of which is the focus of this thesis. The risk of rutting in the AC layers is highest just after construction and then generally diminishes as the materials harden due to traffic and environmental aging. AC rutting has a large impact on life cycle cost because it mostly occurs at the beginning of the life cycle and because failure by rutting often requires removal and replacement of the affected layers or other costly remedial construction, which makes rutting a crucial consideration. Although rutting performance of AC pavements has been characterized by various studies, there has been little study of the effects of AC micromechanical structure on in-situ rutting deformation accumulation mechanisms. In addition, the analysis of the predicted pavement performance variability and the development of a reliability-based design method considering all significant sources of variability have been lacking. The goal of this thesis was to explain the rutting accumulation mechanisms for pavements with conventional and modified asphalt mixes and develop a comprehensive rutting performance prediction procedure considering the fundamental material properties, in-situ deformation accumulation mechanisms and the effects of various sources of variability, based on investigation of the fundamental problems in current test methods, structural models and performance prediction procedures. An innovative method was developed to quantify the precision and bias in repeated simple shear test at constant height (RSST-CH) laboratory test results for specimens with different dimensions and to determine the effects of variability on predicted rutting performance. Specimen size requirements for two different asphalt mix types were proposed based on the results of the analysis. The effects of test temperature and specimen volume on test variability were also investigated. A reliability based rutting performance prediction procedure was developed that considers the variability in laboratory test results, layer thicknesses, stiffnesses, and measured in-situ performance. The effects of input design parameter variability on predicted performance were determined using the calculated distributions of calibration coefficients. By using these calibration coefficient distributions, asphalt layer design thicknesses for different reliability levels can be predicted without performing computationally intensive calculations, such as Monte Carlo simulations, facilitating incorporation of reliability into design software. The general procedure developed for specific tests and AC rutting in this thesis can be applied to other distresses. The use of X-ray computed tomography (CT) images was extended from previous work in a new empirical approach developed to investigate the changes in AC microstructure caused by full-scale accelerated pavement testing with a Heavy Vehicle Simulator (HVS), by using images taken before and after HVS rut tests. A viscoelastic micromechanical finite element model was also developed to investigate effects of binder and aggregate properties on shear resistance using the microstructural model developed from the imaging process with laboratory specimens. The approach was used to investigate the differences in performance under full-scale loading of two mixes, one dense graded with polymer modified binder and the other gap-graded with rubberized binder. It was found that shear related deformation appeared to control the long term rutting performance of the AC pavement layers while densification was primarily an initial contributor at the very early stages of the trafficking. A high concentration of aggregate interlock in the polymer modified mix, as a result of the dense gradation and larger aggregate sizes, appears to have resulted in greater dissipation of shear stresses and therefore greater shear resistance. The lack of this interlocking effect for the rubberized gap-graded mix is proposed to have caused the earlier failure in the full-scale HVS test sections. Important differences in aggregate movement and air-void changes were also observed between different overlay thicknesses indicating the depth of the rut phenomenon, important information for the design of overlays on aged AC as well as for asphalt overlays on concrete pavements. Recommendations are proposed to improve design and construction of asphalt surfaced pavements based on these findings.