Fig. 6
Segmented features from Micro-CT slice used for model geometry generation. (a) Bacterial capsules are identified and shown in green, representing embedded self-healing agents. (b) Cracks are extracted via skeletonization and highlighted in red, indicating tortuous potential transport pathways. (c) Pores are visualized in cyan, showing limited void regions not occupied by capsules or cracks. These segmentations form the basis for constructing a finite element mesh that captures the heterogeneity of the cracked concrete microstructure and informs simulation of healing agent diffusion and precipitation. Note: The limited cyan regions in (c) indicate high capsule/crack density in this section, reducing free pore space. This image represents a localized region with relatively high capsule and crack density, resulting in limited pore space. While this section was chosen to highlight segmentation accuracy, it does not reflect the overall distribution across the specimen. Capsule presence and pore space vary spatially, as observed during image preprocessing.
