Development of a mechanistic model for geogrid reinforced flexible pavements

This dissertation describes the development of a mechanistic response model for the analysis of geogrid base reinforced flexible pavements and to investigate the response benefit of geogrid-reinforcement. This is an axisymmetric finite element structural model that considers the nonlinear, stress dependent pavement foundation geomaterials, i.e., unbound base aggregates and fine grained subgrade soils, anisotropic behavior of the granular base materials, as well as the compaction and preloading induced residual stresses in the granular base. The mechanistic model was used to investigate geogrid reinforcement mechanisms by considering locked-in horizontal residual stresses formed around the geogrid location for a proper modeling of the stiffening effect shown by the geogrid. These residual stresses are taken as initial conditions in pavement response analysis. To validate the response model, full-scale flexible pavement test sections, geogrid reinforced and unreinforced, were constructed at the University of Illinois Advanced Transportation Research and Engineering Laboratory (ATREL). Results from full-scale tests showed reduced critical pavement responses due to the inclusion of geogrid reinforcement. Higher percent reductions in lateral deflections in the aggregate base were realized especially in the direction of traffic. The anisotropic characterization of the granular base was needed to better capture the measured responses from the full-scale test sections. Field forensic and pavement trench studies were used to further calibrate the mechanistic response model. Typical increased horizontal residual stresses at the level of approximately 41 kPa was also needed 100 mm around the geogrid to better match measured pavement responses of the reinforced sections. As a result of the calibration efforts, the predicted responses at different locations in the test sections compared in general well with the field measured responses under different load levels.