Proposição de um sistema de classificação das propriedades de forma de agregados caracterizados com o uso do processamento digital de imagens para a seleção de materiais brasileiros

Abstract

Digital Processing of Images (DPI) is established as a tool for the characterization of aggregate shape properties (shape, angularity and surface texture). Among the main techniques of DPI used in the area of transport infrastructure is the use of the Aggregate Imaging Measurement System 2 (AIMS2). This research aims to propose a classification system of the aggregate properties characterized using AIMS2, contributing to the selection of Brazilian materials during the design process for the production of Hot Mixture Asphalt (HMA) more resistant to major pavement distresses. A database analyzed with AIMS2 was organized with twenty distinct aggregates with mineralogy from different locations, covering eight Brazilian States. Through statistical methods, the limit values of the classification system proposed were calculated from the database. Subsequently, the quality of the measurements of the aggregate shape properties generated using AIMS2 was evaluated through statistical analysis of repeatability and reproducibility. A method was proposed in this research, based on the combination of two approaches, descriptive statistics and inferential statistics. It was used a fonolitic aggregate from a quarry in the Metropolitan Region of Fortaleza (MRF) produced by different crushing processes and analyzed in two equipments. Finally, the impact of the change of the aggregate shape properties through wear at the Micro Deval (MD) in the formation of the mineral skeleton of HMAs was evaluated through the PDI. An aggregate of gneiss origin collected in the MRF was used in the composition of the HMAs. Its shape properties were analyzed in the AIMS2, before and after of the wear on the MD. The mineral skeletons of the HMAs were analyzed in terms of aggregate interlocking with the use of the Image Processing and Analysis System 2 (iPas2) software. The results showed that: an aggregate may be classified differently, depending on the classification adopted; AIMS2 has met the criteria established for repeatability and reproducibility based on the proposed methodology; to assess the impact on the mineral skeleton of HMAs, when the criterion is aggregate interlock, changes in shape properties should not be evaluated alone but along with changes in the particle sizes and in the mix design parameters. The proposed classification system is a tool with great potential to categorize Brazilian aggregates. Finally, there is a contribution to evaluating and selecting aggregates, still in the mix design process, depending on the properties required for the application of the mixtures produced.