Adrian Andriescu

This study evaluates the rutting performance of asphalt binders used in the construction of the Pavement Testing Facility (11 lanes) at the Federal Highway Administration’s Turner-Fairbank Highway Research Center. The binders were either sourced directly from the asphalt plant tank during construction (“tank binders”) or extracted and recovered from asphalt mixtures sampled during construction (“recovered binders”). The researchers performed laboratory tests, including high-temperature performance grade (HTPG), multiple stress creep recovery (MSCR), and frequency sweep tests. Results from the overall binder ranking, based on analyzed outputs from all tests, showed the superior rutting performance of polymer-modified binders, followed by binders recovered from recycling agent (RA)-modified high-reclaimed asphalt pavement (RAP) mixtures, control mixtures, and the no-RAP mixture. The binder rankings based on nonrecoverable creep compliance at 3.2 kPa from the MSCR test and zero shear viscosity (ZSV) from the frequency sweep test correlated most closely with overall rankings, whereas HTPG and Shenoy parameters showed limitations in effectively classifying binders. Given the applied small strains within the linear viscoelastic region during the frequency sweep test, ZSV appeared to be the most promising viscoelastic-based parameter capable of providing robust evaluations of expected rutting resistance comparable to the MSCR test. The study further validated the effectiveness of RA modifications in accommodating elevated RAP contents while maintaining high-temperature performance comparable to control mixtures with lower RAP levels. In addition, recovered binders exhibited significantly higher expected rutting resistance than tank binders, a finding that can potentially be attributed to RAP incorporation and aging effects during production, hauling, and storage.

This article investigates the potential to streamline the double-edge-notched tension (DENT) test for asphalt binders by optimizing the number of replicates and ligament lengths required to reliably evaluate binder strain tolerance and resistance to fatigue cracking at intermediate temperatures. The standard DENT test requires two replicates at each of three ligament lengths: 5, 10, and 15 mm. This rigorous fracture mechanics–based analysis ensures reliable binder strain tolerance determination, independent of specimen size, beyond the linear viscoelastic (LVE) domain. Data from the Federal Highway Administration’s Turner-Fairbank Highway Research Center of 78 binders were analyzed, and they represent various binder sources, formulations, and modifiers, including crumb rubber terminal-blend, terpolymer, styrene-butadiene-styrene, and high polymer modifiers. The crack tip opening displacement (CTOD) was calculated using all three ligament lengths and compared with a simplified version of the analysis using only two ligament lengths (5 and 15 mm), labeled as “CTOD2P,” and a further simplified index called the strain tolerance index (STI), which is based on a single ligament length (10 mm). All parameters were calculated with two sample sets to evaluate repeatability between replicates and the analysis parameters. Statistical analysis (F- and t-tests) showed that CTOD2P and corrected STI are generally statistically indistinguishable from full CTOD for most binders, supporting the feasibility of a reduced protocol for ranking binder cracking resistance beyond the LVE domain. The recommended procedure uses two replicates at a 10-mm ligament length to obtain STI and its coefficient of variation (COV), with a 5 % COV threshold above which a third specimen is advised, and the results further indicate that uncorrected STI can be used directly—without conversion to CTOD—as a practical, standalone index for routine specification, quality control, and binder.

Several highway agencies in various countries are interested in utilizing open-graded friction courses (OGFC) as surface layers for various safety and environmental benefits. Oxidative aging in surface layers contributes to the loss of adhesion between binder and large aggregates, which leads to raveling as the predominant distress in OGFC. The epoxy-modified asphalt (EMA) technology offers an alternative paving material that provides an opportunity to enhance the durability of OGFC. Although the use of epoxy-modified OGFC in New Zealand and the Netherlands has shown superior resistance to raveling, one of the major challenges for wider implementation is the lack of comprehensive understanding of the reactivity between epoxy components and the reactive compounds in base binders. This study investigates the base-binder dependency of epoxy reactivity, effectiveness, and performance of reactive polymer-modified asphalts. To this end, four base-binder sources, carefully selected to represent different chemistry and potential reactivity with epoxy modifiers, were modified with various dosages of the epoxy package (i.e., 0% to 25% by weight), and tested using bending-beam rheometer, the dynamic shear rheometer’s concentric-cylinders geometry, asphalt-binder quality test, and Fourier-transform infrared spectroscopy. The proposed framework showed effectiveness in characterizing EMA reactivity and optimizing epoxy dosages. Results highlighted the limitations of some standard test methods and proposed specification limits when evaluating EMA. This study is part of a larger effort to advance the knowledge and state of practice of EMA technology for successful implementation in the U.S.

This study examined rheological and failure-based approaches for the optimization of two bio-based and two petroleum-based recycling agents (RAs). The results showed that optimizing the RAs based on the performance grade (PG) approach did not fully achieve the rheological and failure characteristics of the target asphalt binder. This was observed for both a bio-based and a petroleum-based RA. The asphalt binder quality test (ABQT) and Glover–Rowe (G-R) approaches showed potential as surrogate methods for estimating the optimum dosage of RAs, whereas the double-edge notched tension (DENT) approach provided a more cost-effective dosage. This needs to be further evaluated through future mixture-level testing and in situ experiments. The selected dosage of each RA, based on the high-end PG, was then applied to a mixture containing 40% reclaimed asphalt pavement (RAP) to achieve the target performance, which was a mixture with 20% RAP. The mixture characteristics were evaluated using the Hamburg wheel tracking test (HWTT) for rutting and moisture damage characterization and the indirect tensile asphalt cracking test (IDEAL-CT) for cracking characterization. The results showed that the addition of RAs improved the cracking resistance of the 40% RAP mixtures without significantly affecting their rutting resistance. The RAs used in this study did not negatively affect the moisture damage resistance of the mixtures. After long-term aging conditions, all mixtures exhibited diminished cracking resistance. The comparison between the aging protocols highlights varied performance among these mixtures, underscoring the necessity for further research in the selection of appropriate aging protocols. Overall, the results revealed that bio-oil-based RAs compare favorably with petroleum-based ones.

The asphalt community is focused on the paradigm shift in mixture design from volumetrics to an optimization procedure based on performance tests called balanced mixture design (BMD). Laboratory performance tests are being assessed for their relationship to binder properties and field performance. However, existing BMD practice does not account for long-term oven aging (LTOA). This gap limits the appropriateness of thresholds and the potential of BMD for improving pavement performance. The Federal Highway Administration’s Asphalt Materials Research Program has launched a three-phase study to provide insights into relevant performance testing. Phase I compared seven cracking performance tests after short-term aging, whereas Phase II evaluated the sensitivity of selected performance tests to LTOA. Phase III, in which seven loose mixtures collected from Montana, Vermont, Ohio, and Florida were subjected to LTOA, is presented in this paper. The indirect tensile cracking, Illinois Flexibility Index, dynamic modulus, and cyclic fatigue tests were employed before and after two proposed LTOA conditioning protocols. Additionally, chemical, rheological, and fracture properties were obtained from extracted binders to obtain a better understanding of the impact of binder properties on mixture performance indices. Results illustrate that LTOA is critical for a comprehensive BMD framework and, thus, some of the testing protocols, indices, and thresholds need to be refined further.

A study was conducted on epoxy-modified asphalt (EMA) diluted with two binders of a different chemical structure, to understand the effect of curing rates and long-term postcuring characteristics. Dynamic shear rheometer test results did not show a clear trend of curing time and dosage percentage impact for the two base binders; parallel-plate geometry may not be well-suited to study EMA because of the thermosetting properties of curing. The asphalt binder test (ABT) method was used to assess the effect on rheology of short-term and long-term curing. Analysis of the ABT data to date suggests that the binder source has a notable effect on the composite properties. The impact of long-term storage was also found to be dependent on the asphalt binder used. Findings from this study were also applied to two field binders; one was modified using styrene-butadiene-styrene polymer, and the other contained epoxy. Curing times were determined using the ABT test. The Dongre Workability Test (DWT) was also conducted on open-graded friction course mixes and the mix curing times were found to be significantly higher than those determined on the binders. A method based on DWT data to determine a “workability window” was also explored. The Strategic Highway Research Program (SHRP) sol-type binder AAG-2 was found to exhibit little to no reactivity with the epoxy modifier used in this study. The SHRP gel-type binder AAD-2, however, was found to be highly reactive. The short-term curing temperature was found to affect the rate of curing for base binder AAD-2. The curing rate was found to be highest at 165°C followed by 120°C, and the rate was at its slowest at 25°C. AAG-2 did not show any onset of curing at either temperature.