Comparison between pore size distributions of the fine aggregate matrix obtained by micro-computed X-ray tomography and a set of linear regressions

The Fine Aggregate Matrix is composed of asphalt binder, fine aggregates, filler materials, and pores. It has been used to understand the behavior of asphalt mixtures since fatigue damage in asphalt mixtures is frequently related to small-scale phenomena with considerable influence off the fine part of the mixture. In addition, several researchers point out the importance of taking into account the pore size distribution in the designs of these samples, as they are related to moisture damage. This research aimed to compare the method of determining pore size distribution of nine Fine Aggregate Matrix samples using X-ray micro-computed tomography to the method of determining pore size distribution by analyzing a set of linear regressions between the pore volume and the density of the samples. The results showed that the pore distribution obtained through linear regression resembles the pore distribution obtained from X-ray computed microtomography and that the average sample density overestimated the linear regression set model by up to 6.1 % of the pore volume. The observed sampling error was less than 0.59 %, which shows that the number of samples was sufficient to estimate the pore volume of the FAM2. Thus, the hypothesis test result confirmed that the difference cannot be considered significant for a 95 % confidence level. Therefore, the linear regression set model can be considered adequate to describe the pore volume obtained by X-ray micro-computed tomography.