Interface Modeling of Geosynthetics Reinforced Asphalt Pavement Structures

Abstract: Geosynthetic reinforcement in asphalt pavements improve pavement performance by enhancing aggregate interlock and reducing lateral movement of pavement layers. They improve compressive strength, stiffness, and reduce permanent deformation in granular base, subbase and subgrade layers. However, factors such as subgrade strength, base thickness, and moisture content significantly influence their effectiveness. When used in Hot Mix Asphalt (HMA), Geosynthetics improve resistance to delamination by reinforcing the concerned layers. Various modeling approaches, including finite element methods (FEM) and artificial neural networks, have been applied to analyze these geosynthetic reinforced pavement structures. Modeling of interfaces between pavement layers in FEM present challenges associated with capturing their bonding conditions, which are crucial to pavement longevity. Poor bonding conditions between pavement layers can result in a significant reduction in service life. Also, debonding mechanism has been a problem in asphalt pavement for a while, which needs to be explored. Though numerous approaches have been explored to evaluate bonding conditions, numerical errors in interface modeling can lead to significant inaccuracies in predicting pavement performance.
To address this research gap, three pavement test sections (Control, Geogrid, Geotextile) were constructed at Rowan University, New Jersey in the year of 2022. These sections were analyzed under heavy vehicle simulator to understand the reinforcing mechanism. Geocell reinforcements were placed in the middle of HMA section. Results indicated the superior performance of geogrid reinforcement over geotextiles. Further, field asphalt cores were collected from these sections to quantify bonding between geosynthetics and HMA materials. The bond strength was assessed through Interlayer Shear Stiffness (ISS) test. In parallel, a bilinear cohesive zone model (CZM) was employed to develop a FEM model for geosynthetic reinforced pavements. After initial geometric calibration using EverStressFE, the FEM model accurately predicted the interface bonding behavior, achieving an RMSE of less than 10% (p-value = 0.1). The developed CZM parameters and framework were used to simulate long term pavement performance. Sensitivity analysis of various design parameters (thickness, modulus and bonding parameters) was also explored simultaneously. It was found that there were significant interactions between design parameters. Subsequently, a prediction model for horizontal tensile strain under HMA layer was developed for various bonding conditions using response surface model.