Laboratory creep and fatigue testing was performed on five Superpave surface hot-mix asphalt mixtures placed at the Virginia Smart Road. Differences in creep and fatigue response attributable to production and compaction methods were investigated. In addition, changes in creep response resulting from differences in specimen size were evaluated. Further, an evaluation of the effects of loading frequency, presence of rest periods, and specimen location within the pavement on fatigue life was conducted. Creep compliance values were determined using viscoelastic-based calculations, and time-temperature superposition was used to generate mastercurves. Reported creep compliance response models from the literature were found inadequate for accurately describing the creep compliance mastercurves generated during this study. Differences in creep response between specimens of different sizes were found to be due to specimen and test variability, rather than size. An evaluation of the effects of laboratory and plant production and laboratory and field compaction was inconclusive as material variability appeared greater than production or compaction variability. Simple regression models were found to be satisfactory for use in the development of prediction models for fatigue, although test data are necessary for calibration to particular mixture types. No relationships were found between fatigue model coefficients and volumetric properties of the mixtures tested because of the limited range of volumetric properties. Variability in volumetric properties between the mixtures produced at the plant and those produced to match the job mix formula did not significantly influence the predicted laboratory fatigue performance. Laboratory fatigue lives were similar between the laboratory-compacted fatigue specimens and specimens cut from the pavement; differences observed in performance were attributable to different air void contents. Predicted fatigue life was found to be statistically independent of the frequency of the applied loads or presence of rest periods for the mixtures, frequencies, and rest periods considered in this study. Minimal differences were observed between fatigue life predictions for plant-produced, field-compacted specimens cut from different locations in the pavement. This study contributes to the understanding of the factors involved in creep and fatigue performance of asphalt mixtures. The mixture responses characterized by this study are related to the rutting and fatigue performance of asphalt pavements. The choice of appropriate asphalt materials to resist rutting and fatigue deterioration will result in reduced maintenance needs and longer service lives for pavements.

The primary objectives of this research study was to characterize properties of two NCDOT Superpave mixes with regards to fatigue distress, and to develop phenomenological fatigue relationships for these mixes based on various levels of strain, asphalt content, air void content, and temperatures. Of particular importance was the sensitivity of the Superpave mixes to asphalt content and air void content that are usually expected in-situ. Fatigue characterization of typical pavement sections using NCDOT fatigue models and using mechanistic analysis procedure suggests that the pavement fatigue life is sensitive to the mix variables and test temperatures considered in this study. A decrease in asphalt content by 0.5-percent (by wt. of mix) results in decrease of 18 to 25-percent fatigue life. An increase in 2% air void content will reduce pavement life by about 40% for 12.5-mm mixes, and by almost 60% for 19-mm mixes. An increase in temperature was found to result in decrease in fatigue life of pavement section under consideration, although, fatigue testing was conducted in controlled-strain mode-of-loading. A 5 & deg;C increase in temperature results in about 25-percent reduction in pavement life. Based on the overall result of the analysis, it appears that 19-mm mix is more sensitive to mix and test variables as compared to the 12.5-mm mix. The study of frequency effect on fatigue life of typical pavement section suggests that prediction of fatigue life of pavement section is independent of load frequency used in fatigue test as long as tensile strain is computed based on the same load frequency.

Laboratory fatigue testing was performed on six Superpave HMA mixtures in use at the Virginia Smart Road. Evaluation of the applied strain and resulting fatigue life was performed to fit regressions to predict the fatigue performance of each mixture. Differences in fatigue performance due to field and laboratory production and compaction methods were investigated. Also, in-situ mixtures were compared to mixtures produced accurately from the job mix formula to determine if changes occurring between the laboratory and batch plant significantly affected fatigue life. Results from the fatigue evaluation allowed verification of several hypotheses related to mixture production and compaction and fatigue performance. It was determined that location within the pavement surface, such as inner or outer wheelpath or center-of-lane, did not significantly affect laboratory fatigue test results, although the location will have significant effects on in-situ fatigue life. Also the orientation of samples cut from an in-situ pavement (parallel or perpendicular to the direction of traffic) had only a minor effect on the laboratory fatigue life, because the variability inherent in the pavement due to material variability is greater than the variability induced by compaction. Fatigue life of laboratory-compacted samples was found to be greater than fatigue life of field-compacted samples; additionally, the variability of the laboratory compacted mixture was found to be less than that of the field-compacted samples. However, it was also found that batch-plant production significantly reduces specimen variability as compared to small-batch laboratory production when the same laboratory compaction is used on both specimen sets. Finally, for Smart Road mixtures produced according to the job mix formula, the use of polymer-modified binder or stone matrix asphalt was shown to increase the expected fatigue life. However, results for all mixes indicated that fatigue resistance rankings might change depending on the applied strain level. This study contributes to the understanding of the factors involved in fatigue performance of asphalt mixtures. Considering that approximately 95% of Virginia's interstate and primary roadways incorporate asphalt surface mixtures, and that fatigue is a leading cause of deterioration, gains in the understanding of fatigue processes and prevention have great potential payoff by improving both the mixture and pavement design practices.

The final product of the Strategic Highway Research Program (SHRP) Asphalt Research Program is the SUPERPAVE (registered trademark) mix design system for new construction and overlays. This system employs a series of new performance-based specifications, test methods, and practices for material selection, accelerated performance testing, and mix design. This report documents these new specifications and procedures in a format suitable for eventual American Association of State Highway and Transportation Officials (AASHTO) standardization.