Análise microestrutural de um pavimento asfáltico do município de Alegrete através do processamento digital de imagens

Abstract

This study aims to determine the feasibility of using digital image processing to analyze the granulometry, orientation and distribution of aggregates in samples of asphalt mixtures. The specimens used in this research were obtained in two ways, 4 samples were molded in the field using the Marshall compaction method and the other 4 were extracted with the aid of a rotary probe, from an asphalt pavement compacted by tandem and pneumatic rollers. Such samples were taken to the laboratory, where they were identified, sectioned into thin slices and scanned in a multifunctional printer, where the sections were digitized. After that, the images were processed with the help of ImageJ software and the area properties, shape description, area fraction, centroid, and Feret diameter of each of the particles were obtained, these data were organized and the results processed. As a result, the analysis of the granulometry of the images of the samples showed that the molded ones presented very similar granulometries in all their slices, whereas the extracted ones, presented more distinct granulometries between their slices and with unequal distribution, however, both samples had granulometric curves that were not coincident with the design curve. Both types of samples showed similar distribution of orientations, with no direct influence of the compaction method on these parameters. The results also showed that segregation may have occurred in the execution of the asphalt pavement from which the samples were extracted, but due to uncertain granulometric properties, this fact was considered inconclusive. However, it is concluded that digital image processing is a promising method, which can become effective for the analysis of properties of asphalt mixtures, but it needs some improvements to obtain even more significant results and achieve this goal.