Studying rutting performance of Superpave asphalt mixtures using unconfined dynamic creep and simple performance tests

Abstract

This study aimed at evaluating asphalt mixtures for rutting using two test procedures. The first procedure was the dynamic creep test, which was performed using the 5-kN Pneumatic Universal Testing Machine (UTM-5P). The second procedure was the flow number test which was performed by the Superpave Simple Performance Tester (SPT), currently known as the Asphalt Mixture Performance Tester (AMPT). Test specimens were prepared using crushed limestone aggregate and 60/70-penetration-grade asphalt binder having a performance grade (PG) of 64-10. Two aggregate gradations were used and compared in this study: a gradation passing above the restricted zone (ARZ) representing a fine gradation, and another gradation passing below the restricted zone (BRZ) representing a coarse gradation. Asphalt mixtures were compacted using the Superpave Gyratory Compactor (SGC). The SGC samples were then cored and sawed to produce fabricated test specimens of the desired height (150 mm) and diameter (100 mm). Specimens were tested at four temperature levels: 40°C, 50°C, 55°C, and 60°C. For comparison purposes, identical test parameters were used for both tests including: specimen dimensions, load frequency, load and rest periods, contact stress, and deviator stress. Test results were analysed to investigate the permanent deformation behaviour of the asphalt mixture with the temperature. The results for fine-graded mixtures and coarse-graded mixtures were analysed and compared. A comparison between the two test procedures was made based on the test results. The analysis and comparison were made based on the number of cycles at failure, the strain at failure, the number of cycles to reach 1.5% strain, and the strain at 1000 cycles. Both one-way ANOVA and two-way ANOVA procedures were used in the comparison. Results showed that a significant difference between ARZ asphalt mixtures and BRZ asphalt mixtures in the measured properties existed. The significance level was found to be strongly related to the test temperature. Results also showed that the flow number test and the dynamic creep test results had different behaviour with test temperature and sometimes opposite behaviour. The significance of the difference was also found to have an interaction with the test temperature. Based on the results and the comparison, it was clear that the SPT flow number test showed better accuracy and reproducibility of test results. The flow number test results also showed better fitting and no departures from the expected trend, unlike the dynamic creep test results.