Analysis of the evaluation indices from TSRST

Because the thermal stress restrained specimen test (TSRST) is recommended by the Strategic Highway Research Program (SHRP) as a test method to evaluate low-temperature cracking resistance of asphalt mixtures, many studies have been carried out to investigate the performance of asphalt mixture with this test using fracture temperature as the indicator. However, few of them focus on discussing its reliability.In this paper, six kinds of asphalts mixtures are evaluated by TSRST,and it is found that the evaluation results obtained by the four indices are contradictory. It is also inferred from the results of eachindex coefficient of variation that fracture temperature and transition temperature are more stable when they are used to evaluate low-temperature performance of asphalt mixtures. Through the application of principal-component analysis (PCA) and the Boston Consulting Group's matrix on the TSRST results, it is demonstrated that fracture temperature used as the indicator of low-temperature performance of asphalt mixture is reliable, and this conclusion is verified by gray relationanalysis between the four indices from TSRST and bending-strain-energy density.