Image-Based Quantitative Pore Analysis of Cementitious Materials: A Scale-Resolved Review of X-Ray CT Methods and Metrics

Abstract

X-ray computed tomography (CT) has become the reference non-destructive technique for the three-dimensional characterization of the pore structures of cementitious materials. This review maps 140 application studies—spanning cement paste, mortar and concrete, from air-entrained, cellular and recycled-aggregate systems to alkali-activated binders, ultra-high-performance concrete (UHPC) and 3D-printed materials—and the underlying image-analysis methodology is inventoried separately and excluded from all frequency statistics. Each study is classified by voxel-size class (sub-micron, <1 µm; micro, 1–20 µm; meso, 20–100 µm; macro, >100 µm; multiscale), with voxel size being a sampling parameter rather than a spatial resolution, and by the pore metrics it reports under six metric families. The synthesis reveals similar concentrations at the micro (48/140) and meso (50/140) scales, a dominance of porosity and pore-size distribution at every scale, and pronounced under-use of higher-order shape and spatial statistics: flatness, elongation and pair-correlation functions each appear in less than 2% of studies. Systematic discrepancies between scales are analyzed: resolution-conditioned porosity, the representative elementary volume–resolution dilemma, segmentation-induced variance comparable to material effects, and the persistent CT–mercury intrusion porosity mismatch. Priorities are standardized segmentation benchmarks, scale-bridging registration protocols, routine uncertainty reporting and broader adoption of quantitative shape and spatial descriptors. Provisional assignments are flagged in the dataset.