A test of the resistance to vertical segregation in self-compacting concrete using image processing techniques

This paper presents a novel approach, utilizing image analysis to accurately and efficiently assess the resistance of self-compacting concrete (SCC) to vertical segregation. Examining varying levels of fluidity across five unique formulations enhanced the investigation of SCC segregation. We obtained images of SCC samples using a multifunction scanner, which ensured uniformity and reproducibility, thereby establishing a credible foundation for the suggested testing methods. The segmentation process consists of two stages: initial manual segmentation using AutoCAD to establish a critical reference, followed by automated segmentation with ImageJ. This automated approach extracts essential quantitative data, such as paste area and aggregate distribution, while also identifying potential mix defects. Consequently, we presented the longitudinal profiles of aggregate and paste distribution, which allowed us to link formulation parameters to structural outcomes. The study also introduced the Segregation Resistance Indicator (SRI) to evaluate the degree of segregation, classifying the data into three distinct levels for effective comparison between samples. The results demonstrate a robust association with another prevalent segregation indicator, indicating the high reliability of the new methodology. This framework facilitates the establishment of objective criteria for the acceptance or rejection of specific mixtures, thus improving its applicability in various construction scenarios.