Developement of a framework for asphalt performance-related specification and performance prediction
Abstract
Traditional asphalt mixtures have generally involved relatively simple combinations of virgin asphalt binder and aggregates to meet load-bearing needs of the roads and surfaces. Accordingly, simple tests such as Marshall Stability and Flow were used in an effective manner for asphalt mixture screening and quality control purposes. In recent years, there has been a proliferation of asphalt ingredients available to designers, especially in the case of recycled materials, compaction aides, and mixture performance and/or sustainability promoting products. These include reclaimed asphalt pavement (RAP), recycled asphalt shingles (RAS), warm mix agents, antistripping agents, rejuvenators, ground tire rubber, and even waste plastic. These modern, heterogeneous asphalt mixtures exhibit more complex behavior as compared to earlier mixes containing fewer ingredients and predominantly virgin materials. As a result, recent asphalt mixes require advanced performance tests to account for these complexities, while factoring in traffic and environmental loads for the given mixture type being designed. In this study, various existing roads including good and bad performing sections were selected in Illinois and Missouri and after conducting site visits. In field investigations, the main distresses on the Tollway in northern Illinois and across the state of Missouri were identified. Also, several field cores were obtained from mainline and shoulder sections to evaluate the laboratory performance of existing asphalt mixtures across a range in-situ aging levels. Analyzing the available field performance data such as international roughness index (IRI), condition rating system (CRS), and rut depths and comparing them with laboratory testing results provided a robust data set to establish updated performance test thresholds for the Tollway mixture design specification. According to recent literature, mixture performance can be evaluated using various tests to mitigate different distress types such as cracking, rutting, and moisture damage. In this study, fourteen different mixtures produced in 2018 on mainline and shoulders sections across the Tollway system, and four mixtures in Missouri were selected to characterize performance testing trends and to study the ability of the different performance tests to predict pavement performance. To this end, performance tests such as the DC(T), I-FIT, IDEAL-CT, IDT, Hamburg, and TSR were conducted on the collected plant produced mixtures. The process of sample fabrication, ease of conditioning and testing, repeatability and ability to correctly rank various Tollway mix types was taken into consideration in selecting the appropriate performance tests to be used in the Tollway's mix design asphalt specification. The DC(T) test was found to possess the best correlation to field performance, and significantly outperformed the I-FIT test in terms of test repeatability. Both the I-FIT and IDEAL tests returned failing results for a number of SMA mixes, and dense-graded mixes with high recycling content, which have traditionally performed well on the Tollway. This provided additional motivation to retain the DC(T) test as the cracking test to be used in the Tollway's asphalt mixture design specification. The analysis presented in this study, in conjunction with field observations, led to the identification of various cracking types as the primary distresses observed on Tollway mainline and shoulder sections surfaced with asphalt. Rutting and stripping were not found on Tollway asphalt surfaces, nor in Missouri at the present time. The Disk-shaped Compact Tension (DC(T)) test was chosen to be retained in the performance related specification (PRS) for the design of crack-resistant mixtures due to its high degree of correlation with field results and best repeatability. A systematic approach was developed, which allowed different reliability levels to be addressed in the specification, along with a consensus step to take advantage of local practitioner experience. Similarly, for high-temperature performance, Hamburg thresholds for binder course mixtures were tailored for different mixture types and use cases. In some cases, by relaxing Hamburg requirements, designers have more leeway in building crack resistance into the mixture and/or to utilize higher amounts of recycled materials. For SMA mixtures, it was observed that low rut depth mixes were sometimes identified as having stripping potential in the Iowa method. However, similar mixtures have not exhibited stripping in the field. As a result, it is recommended that SMA mixtures with rut depths less than or equal to 4.0 mm after 20,000 passes should be characterized as non-stripping and do not need to be checked for stripping potential through the Iowa method. Based on experimental results, it is also recommended to use the Hamburg test in lieu of the TSR stripping test for moisture sensitivity evaluation. In the event of failing results, the TSR test can be used as a secondary method to assess adequate moisture resistance. Using the specification for both cracking and rutting, a novel grading system that takes into account the performance of the asphalt mixture was developed and the Tollway mixtures were graded based on this system. Also, the characterization of the viscoelastic behavior at low temperature through the DC(T) creep test led to validating the performance of the mixtures that could pass the cracking requirements.
Degree
Ph. D.