Hot mix asphalt time-temperature shifting and fitting techniques: A comparative study

In this study hot mix asphalt (HMA) creep compliance master curves were constructed using different temperature shifting techniques and fitting functions. The study aimed to compare the capability of using five different shifting techniques (numerical, log-linear, William-Landel-Ferry (WLF), Modified Kaelble (MK) and Arrhenius) and two different fitting functions (the Sigmoidal fitting function, and Prony series) to construct creep compliance master curves. To achieve this aim, one hundred and twenty (120) HMA specimens were prepared at three air voids levels (3.4, 4.9 and 6.4%). Half of the prepared HMA specimens were modified using Elvaloy Reactive Elastomeric Terpolymer (Elvaloy® RET at 2% by weight) while the other half remain unmodified. Also, half of the prepared HMA specimens were exposed to short-term aging (STA), while the other half exposed to both short term aging and level 2 long-term aging (LTA2). All the prepared HMA specimens were exposed to a dynamic creep test using a pulse load of 107 kPa at five different testing temperatures (5, 15, 25, 45 and 60 °C). Both graphical and statistical comparisons were made among different shifting techniques and different fitting functions. Generally, The Sigmoidal fitting function was found to be the most accurate fitting function. Comparing with the numerical shifting technique, the best fit between measured and predicted shift factors was found for WLF followed by log-linear, MK and Arrhenius techniques, respectively.