Synchrotron X-ray nanoimaging (XNI), X-ray microscopy (XRM), and CT-based 3D reconstruction have been utilized to reveal nanoscale-to-microscale pore structures of cementitious composites.

Past studies demonstrated the evolution of C-S-H and portlandite around unhydrated clinker phases and quantified pore connectivity and coordination number using skeleton and RDF analysis.

These achievements provide a robust experimental foundation for generating high-resolution datasets required for training deep learning models.

Developed image-processing pipelines using GANs and LIIF for super-resolution of wide-field SEM micrographs, bridging the resolution gap between nano- and micro-scale observations.

Applied U-Net and 2.5D segmentation models to complex pore systems (e.g., waste glass wool-incorporated composites), enabling accurate classification of multiscale pore structures.

These AI-enhanced datasets improve the precision of pore size distribution, connectivity, and tortuosity quantification, which are critical descriptors for mechanical behavior prediction.

Previous experimental works on CO₂-curable cementitious materials, fiber-reinforced composites, and novel binders (CSC, LC³, waste-derived additives) have revealed direct correlations between pore characteristics and compressive/flexural strength.

By integrating these datasets with deep learning regression models, the proposed research aims to establish a predictive framework that quantitatively links pore-scale features to macroscopic mechanical performance.

This methodology will enable the development of a generalizable interpretation model applicable to diverse next-generation cementitious composites.

Incorporation of nanomaterials such as nano-silica, nano-alumina, and graphene oxide to accelerate hydration reactions and refine pore structures.

Demonstrated reduction of calcium hydroxide content and enhancement of C-S-H gel formation, leading to denser microstructures and improved durability.

Utilization of advanced characterization techniques (synchrotron X-ray nanoimaging, SEM, TEM, and 3D CT analysis) to reveal nanoscale-to-microscale changes induced by nanomaterial addition.

Quantification of pore structure evolution, coordination number, and interfacial transition zone (ITZ) improvements, providing fundamental insights into structure–performance relationships.

Nanomaterial incorporation shown to significantly improve compressive and flexural strength, toughness, and resistance to thermal and chemical attacks.

Development of multifunctional cementitious composites (e.g., self-sensing, fire-resistant, and CO₂-curable systems) that leverage nanomaterial properties for next-generation construction applications.

Partial replacement of clinker with calcined clays, limestone, and waste-derived supplementary cementitious materials (SCMs) such as glass wool, waste ceramics, and biomass ash.

Optimization of clinker–SCM blending ratios to balance strength development, hydration kinetics, and CO₂ reduction potential.

Advanced imaging techniques (synchrotron X-ray nanoimaging, SEM, 3D X-ray microscopy) employed to capture hydration product evolution and pore network formation.

Deep learning-based image processing (U-Net segmentation, LIIF super-resolution, GAN reconstruction) applied to quantify pore size distribution, connectivity, and tortuosity across multiple scales.

Machine learning models developed to correlate pore descriptors with macroscopic properties such as compressive strength, carbonation resistance, and chloride penetration.

Establishment of a predictive tool enabling performance evaluation of low-CO₂ cements without extensive long-term experimental testing, enhancing scalability and industrial applicability.

Development of eco-friendly concrete mixtures by partially replacing natural aggregates and cement with waste glass wool fibers and recycled concrete aggregates (RCA).

Reduction of construction waste disposal and CO₂ emissions while promoting circular economy practices in the construction industry.

Investigation of hydration reactions and pore structure evolution in waste-incorporated systems using advanced imaging techniques (X-ray nanoimaging, SEM, and CT-based 3D analysis).

Quantification of porosity, pore connectivity, and interfacial transition zone (ITZ) improvements to evaluate durability and long-term performance.

Evaluation of compressive and flexural strength, chloride penetration resistance, and carbonation resistance in recycled-material concrete.

Demonstrated feasibility of producing high-performance, sustainable concrete with comparable or superior mechanical properties to conventional systems.