An Investigation of the Applicability of Schapery′s Work Potential Model for Characterization of Asphalt Fatigue Behavior

摘要

This paper presents an investigation of Schapery′s work potential model applied to asphalt concrete characterization. Six asphalt concrete mixtures, all consisting of one and the same aggregate and particle size distribution, manufactured using unmodified and SBS modified binders were used. An elastic-viscoelastic correspondence principle and the timetemperature superposition principle are utilized to obtain material characteristic functions, C1(S1), using uniaxial monotonic (constantstrain rate) and cyclic (fatigue) tests at 0, 10 and 20C. The results indicate that consistent and reliable results are obtained for the mixtures investigated using monotonic testing, provided that the applied strain rate is not too low at a given temperature. It was also indicated that the timetemperature superposition principle is applicable as a single material function can be obtained at the reference temperature chosen. Results similar to those obtained at monotonic tests were also obtained at cyclic testing. However, based on statistical analysis, it was noticed that cyclically established C1(S1) curves often differ slightly between different temperatures even when shifted to the reference temperature. This result is probably due to the more complicated excitation history applied during fatigue tests. However, even though it is possible to obtain characteristic material functions from both types of tests, it was not possible to show that monotonic and cyclic testing lead to similar results, i.e. more or less identical C1(S1) curves. It was also investigated, using thermocouples, whether, and if so, to what extent hysteretic selfheating influences fatigue test results. In general, a significant influence was indicated at 10 and 20C, and especially for high amplitude tests (low cycle fatigue tests). Even though several questions still are to be answered, the model based on the work potential theory remains interesting, and is possibly, the most versatile damage model for asphalt characterization available today.