Phase segmentation of concrete x-ray tomographic images at meso-scale: Validation with neutron tomography. (April 2018)

• Record Type: Journal Article
• Authors: Stamati, Olga; Roubin, Emmanuel; Andò, Edward; Malecot, Yann
• Is Part Of: Cement & concrete composites. Volume 88(2018)
• Page Start: 8
• Page End: 16
• Publication Date: 2018-04
• Subjects: X-ray tomography -- Neutron tomography -- Cementitious materials -- Concrete -- Aggregates -- Porosity -- Morphological description -- Image analysis tools -- Phase segmentation; Composite-reinforced concrete -- Periodicals; Concrete -- Periodicals; Composite materials -- Periodicals; Composites de ciment -- Périodiques; Béton -- Périodiques; Composites -- Périodiques; Béton léger -- Périodiques; Cement composites; Composite materials; Composite-reinforced concrete; Concrete; Lightweight concrete; Periodicals; Electronic journals; 620.135
• DOI: 10.1016/j.cemconcomp.2017.12.011
• Languages: English
• ISSNs: 0958-9465

Abstract

Abstract: This work makes the link between the experimental technique of x-ray tomography and the reliable representation of the heterogeneities of a cementitious material at the meso-scale (aggregates, mortar matrix and macro-porosity). An image analysis procedure and its validation are presented. Images coming from x-rays scans are treated to finally obtain a trinary 3D image, where each phase of the material's meso-structure is separated. The separation of the aggregates from the mortar matrix is the technical challenge, due to the physics of x-ray interaction with matter, addressed on this work. Validation of the proposed method is made by comparing the grain size distribution curves computed from images of the same set of aggregates coming from x-rays and neutron scans. The reliably obtained morphology can be used as a basis for direct physical simulations, as well as data to improve morphological models of any type.