Structural Characteristics of Unbound Aggregate Bases to Meet AASHTO 2002 Design Requirements: Interim Report

This report gives the results of a study of the properties of unbound aggregate base materials using both laboratory testing data from full scale field tests in Illinois, Georgia, and Texas, and a model of cross-anisotropic elastic materials to characterize the behavior of the base materials under traffic loads. Using the cross-anisotropic model, the stress distribution in a base course is more realistic than that developed when the aggregate base is considered to be linear and isotropic. The stress distribution based on cross-anisotropic analysis is not only more correct, but it is also more favorable to the unbound aggregate in that significant tensile stresses are found not to occur. The analogy is presented in this report that the response of the aggregate base to the load is as if the stress distribution directly under the wheel load due to anisotropy acts as a moving column under the wheel in which the aggregate essentially produces its own confinement and does not enter into tension. Other findings in this report include the following: 1) The unbound aggregate base material should be modeled as non-linear and cross-anisotropic to account for stress-sensitivity and the significant differences between vertical and horizontal moduli and Poisson’s ratios. 2) The ICAR laboratory testing protocol is efficient and precise and should be considered as a candidate to model the unbound aggregate base. The protocol uses three stress regimes and ten stress levels within each regime to determine stress sensitivity and cross-anisotropy. A system identification method is used to select the five material properties based on the tests results necessary to properly characterize the aggregate base and to satisfy the requirements of elastic work potential theory. 3) The Fast Industrial Process Controls cell is efficient and should be used to characterize unbound aggregate bases. The ratio of the diameter to the specimen height is 1:1. While testing of such sample sizes is discouraged in the literature, improvements made to the IPC cell minimize frictional development between the sample and loading platens resulting in minimal constraint at the sample ends. This is verified in the report based on comparative triaxial testing and finite element analysis. 4) The ICAR testing protocol is an excellent tool for both unbound aggregate characterization and comparative analysis of materials. A compaction study on two very different aggregates (uncrushed river gravel and crushed limestone) was performed in which the aggregates were subjected to impact, kneading gyratory compaction. The difference in the tendency of the compaction techniques to produce varying levels of particle orientation (which affects anisotropy) was evident in the degree of anisotropy measured.