| In this dissertation, the characterization of hot mix asphalt shear resistance and a correlation with rutting performance was studied. The research plan was conducted in two phases. The initial phase was development of a Gyratory Load Cell Plate Assembly to measure shear resistance of hot asphalt mixes. The device was tested on a number of mixtures to develop an analysis methodology for shear resistance and also to verify its ability to distinguish between different mixes in terms of shear resistance. The device successfully differentiated between mixes and showed high sensitivity to variations in mix volumetric components. The second phase concentrated on development of a rutting test methodology and a testing plan to develop correlations between shear resistance and rutting performance. A designed experiment using the unconfined rutting setup was run on 72 mix specimens with different asphalt content, compaction energy and fine aggregate content. The test specimens were adjusted to an appropriate aspect ratio by coring them from 150 mm standard Superpave gyratory compactor specimens. End treatments were performed to reduce end-effects due to surface friction. The rutting test parameters, i.e., deviatoric stress and temperature, were selected based on previous MnRoad test results to achieve a realistic representation of field conditions.; Data analyses were then performed to construct correlations between shear resistance and rutting performance parameters, i.e., flow number, resilient modulus, resilient strain and permanent strain rate at a specific number of cycles. The results indicated that correlations between shear resistance and rutting performance are more significant for mixes with similar gradations. When comparing mixes with different gradations, the correlations are not as good. In addition, shear resistance showed higher variability than any of the rutting performance parameters measured under unconfined conditions. The unconfined setup resulted in opposite trends between shear resistance and rutting parameters for mixes with high fines content as opposed to somewhat courser graded mixes. Based on this research, it is recommended to repeat the same experimental program with a confined setup and seek similar correlations between the two performance parameters. |