University research sheds light on how geogrids improve pavement performance.

Since 2003, researchers at the University of Illinois’ Advanced Transportation Research and Engineering Laboratory (ATREL) have focused on developing a mechanistic-empirical (M-E) design procedure for geogrid-reinforced flexible pavements for low- to moderate-volume roads. Their ultimate goal is to provide project owners and designers with more advanced ways of designing and analyzing pavement structures.

Improvement to the pavement systems provided by geogrid reinforcement is frequently quantified by traffic benefit ratios (TBRs), traffic improvement factors (TIFs), and base course reduction (BCR) based on direct comparisons of the performance of reinforced sections with identical unreinforced sections.

Current design methods, including the American Association of State Highway and Transportation Officials (AASHTO) PP 46-01, offer a convenient method for designing geogrid-reinforced pavements, but they are limited and do not fully account for the reinforcement benefit in various design conditions. And, as noted in a National Cooperative Highway Research Program (NCHRP) project report (Christopher et al., 2001), industry professionals remain troubled by the lack of evidence for the mechanisms contributing to base reinforcement, as well as the absence of cost-benefit data and appropriate design methods for facilitating bidding and construction.

In a mechanically stabilized layer, aggregate particles interlock within the geogrid and are confined within the apertures, creating an enhanced composite material with improved performance characteristics. The structural properties of the mechanically stabilized layer are influenced by the magnitude and depth of the confined zones.

Tensar International Corporation decided to respond to these lingering concerns by supporting carefully controlled laboratory and field tests at ATREL. The company was confident that the research conducted by the University of Illinois would generate volumes of valuable and precise pavement response data and advance the profession’s understanding of how and why geogrids improve performance in flexible pavements, while also beginning the process of moving the design community away from purely empirical approaches.

Even before ATREL began its work, there were compelling reasons for transitioning to a more advanced mechanistic-based design methodology. Evidence was mounting that the transportation industry would soon require a new, more efficient design approach given the need to repair or replace large percentages of existing low- to moderate-volume infrastructure, the growing scarcity of easily accessible aggregate, and the rapid increase in the cost of asphalt, fuel, and other petroleum-based products. If widely implemented and correctly constructed, geogrid base reinforcement could provide a practical means of addressing these challenges.

Research approach
Research at ATREL essentially fell into two phases: developing a mechanistic model, and full-scale testing for mechanistic model validation and for distress model developments.

Phase 1: Mechanistic model development—A mechanistic model, written in Fortran programming language, was initially developed for the analysis of geogrid-reinforced flexible pavements based on the finite element approach. To properly account for the nonlinear, "stress-dependent" resilient behavior of granular materials and fine-grained soils under repetitive loading, a variety of nonlinear material models were used for characterization of the granular base and subgrade soil layers. In addition, the developed model also took into account the directional dependency (anisotropic modulus properties) of the granular materials in the mechanistic analysis for accurately predicting critical pavement resilient responses (Kwon et al., 2005).

Itasca Consulting Group, Inc., investigated aggregate and geogrid interactions using the discrete element modeling (DEM) technique. The analysis indicated that a "stiffened zone" developed on both sides of the geogrid during compaction and traffic loading due to the geogrid-aggregate interlock. As a result of increased contact forces and stresses around the geogrid, the stiffness of the adjacent unbound aggregate increased significantly and improved overall pavement performance. These investigations demonstrated that residual stress and confinement effects must be considered in numerical analysis. The findings of DEM studies were accounted for in the developed mechanistic model by assigning initial horizontal residual stress around the geogrid (Kwon et al., 2008).

This discrete element rolling wheel model simulates stress transfer from an imposed load on an asphalt-aggregate subgrade pavement section.

Phase 2: Full-scale testing—Testing of the full-scale pavement sections was conducted using the mobile accelerated testing loading assembly at ATREL. Nine instrumented sections were tested to quantify the effectiveness of geogrid-reinforced flexible pavements under a moving wheel load. Accelerated pavement testing (APT) offered an attractive method of compressing many years of vehicular traffic into a relatively short period. The full-scale APT was essential for generating reliable data and truly understanding how a geogrid-reinforced pavement performs under real-world conditions.

The APT response testing was carried out using the following variables: tire loading, tire pressure, speed, and tire configuration. The response test results were used to validate and calibrate the mechanistic model. Measured responses under different load levels indicated that the geogrid reinforcement had reduced pavement responses such as tensile strains at the bottom of the hot mix asphalt, vertical pressure and strains on top of the subgrade, and lateral movements in the aggregate base layer (Al-Qadi et al., 2008).
Further, the unreinforced pavement sections resulted in the highest surface rutting, and the pavement failure due to subgrade shear was more pronounced in the unreinforced control sections. The APT trafficking test results are currently undergoing evaluation to develop distress models for predicting the rutting and fatigue performance of flexible pavement reinforced with Tensar geogrids.

These results enabled Tensar International to develop an Advanced Base Course Reinforcement software application that uses TBRs in combination with enhanced layer coefficients to explain how geogrids stiffen the aggregate layer above the geogrid and how retained stiffness, along with damage reduction, enhances overall pavement performance. While this software application is an interim step to a full M-E based design approach, it accounts for the confinement benefits of geogrids by using both improved pavement response and increased resistance to damage.

Work to date means industry professionals are no longer limited to designing flexible pavements using only AASHTO 1993 with TBRs, but more research is needed to further validate the M-E approach. In particular, we need to understand how residual stresses develop and lock-in around the geogrid reinforcement as a function of its location in the base course. We also need to further explore different types, properties, and configurations of geogrid and their interaction with unbound aggregate types, size, and shapes, as well as the construction and subsequent trafficking loads and patterns experienced by the pavement section. In doing so, this work will lead to development of optimal geogrid configurations and geometries, and allow us to maximize pavement performance and economics.

Final thoughts
The transportation industry is facing big challenges, but M-E design methods offer practical solutions for designing more reliable pavements that last longer and are easier to maintain. We now have a framework for understanding the role of mechanically stiffened base course materials in reinforced pavement structures.

The research highlighted here suggests that properly designed flexible pavement structures with geogrid base reinforcement can significantly enhance pavement performance. Experience indicates that geogrid-reinforced, flexible designs are faster to construct and are less susceptible to the kind of weather-induced construction delays that can bring lime-stabilized, thick concrete or unreinforced, flexible pavement installations to a halt. Equally important, maintenance work is easier, less disruptive, and more affordable since full-depth rehabilitations can be minimized.

Adopting M-E design methods provides an opportunity for project owners to build more efficiently by using less aggregate in their designs. For projects to be successful, however, the industry needs to adopt specifications that de-emphasize the physical characteristics of geogrid products (such as a specific aperture size). This means placing more focus on how the geogrid-reinforced layer (the "composite layer" of geogrid-reinforced base course material) performs and behaves in the pavement. Current practices are weighted to try to understand how slight changes in geogrid material properties affect performance. Moving to performance-based specifications poses a number of challenges, including how to understand and predict a pavement’s performance based on the physical properties of the geogrid reinforcement.

Joe Cavanaugh, P.E., is vice president of technology for Tensar International Corporation. He has 14 years of experience in geosynthetics and geotechnical design and construction. Contact him at jcavanaugh@tensarcorp.com. Jayhyun Kwon, Ph.D., is a senior pavement engineer for Tensar International Corporation. Contact him at jkwon@tensarcorp.com.

SIDEBAR
References

• Al-Qadi, I.L., S. Dessouky, J. Kwon, and E. Tutumluer, 2008 (expected), Geogrid in Flexible Pavements: Validated Mechanism Presented at the 87th Annual Meeting of Transportation Research Board (TRB), accepted for publication in the Transportation Research Record.
• Christopher, B.R., Berg, R.R., and Perkins, S.W., 2001, Geosynthetic Reinforcements in Roadway Sections, NCHRP Synthesis for NCHRP Project 20-7, Task 112, Transportation Research Board, Washington, D.C.
• Kwon, J., Tutumluer, E., and Kim, M., 2005, Development of a Mechanistic Model for Geogrid Reinforced Flexible Pavements, Geosynthetics International, 12:6, pages 310-320.
• Kwon, J., Tutumluer, E., and Konietzky, H., 2008, Aggregate Base Residual Stresses Affecting Geogrid Reinforced Flexible Pavement Response, International Journal of Pavement Engineering, Volume 9, Issue 4, pages 275-285.
• Kwon, J., Tutumluer, E., Al-Qadi, I.L., and Anochie-Boateng, J., 2007, Geomaterial Characterizations of Full-Scale Pavement Test Sections for Mechanistic Analysis and Design, ASCE Geo-Institute, Geo-Denver 2007 Conference, Denver.

SIDEBAR 2
Measuring performance of two roadway structures
Data was collected from two projects with sections of reinforced and unreinforced pavement. The structures on Palomar Street and Industrial Boulevard were installed in November 2003 during reconstruction of two high-traffic roadways in Chula Vista, Calif. During construction, a 50-year-old steel water line was found to be much shallower than anticipated. Realigning the water line would have added $250,000 to the project.

Tensar International Corporation recommended using biaxial geogrids for base reinforcement over the sections of the utility line that were too shallow for conventional construction. Tensar’s design criteria predicted that the geogrids with 6 inches of crushed aggregate base and 11 inches of asphalt concrete could be used to reduce pavement thickness by 11 inches on Palomar Street and 8 inches on Industrial Boulevard without any sacrifice in strength or performance.

In 2008, Tensar used KESSLER dynamic cone penetrometer readings to measure the in-situ California bearing ratio of the aggregate base at six locations. Analysis revealed that the unreinforced sections required 10 to 13 inches of additional aggregate to achieve a level of stiffness comparable to the 6-inch sections reinforced with geogrids (see Table 1). It also confirmed the stiffness retention within the aggregate base course of the reinforced sections even after five years of traffic volume.

A linear elastic mechanistic-empirical analysis also revealed that the 6-inch reinforced section was performing at a higher calculated traffic volume, as measured in equivalent single axle loads (ESALs). Based on the recorded stiffness from the field for each of the aggregate base sections, the reinforced sections could have been reduced to 9.5-inches of asphalt concrete in Palomar Street and 8.5-inches in Industrial Avenue while still providing the same level of performance as the unreinforced sections (Table 2).