The pore radius propagation approach: computational characterization of pore structures in granular media

Accurate quantification of pore structures and characterization of pore-scale fluid behavior are fundamental to the modeling of unsaturated flow and multiphase interactions in geotechnical systems. This study presents a computationally efficient Pore Radius Propagation (PRP) approach for estimating pore-size distributions from computed tomography (CT) datasets. The proposed method was validated using open-access in-situ CT images of Hamburg sand and glass-bead specimens subjected to unconfined compression and drainage-imbibition cycles. The PRP approach characterized the microstructural features of both materials and enabled the derivation of corresponding water retention behavior. Furthermore, it successfully captured the spatial evolution of pore structures induced by irregular particle rearrangement and the stochastic formation and collapse of localized pores during uniaxial loading. The proposed framework also accurately reproduced the microscale spatial distribution of capillary water under specified suction conditions, demonstrating its capability to represent pore-scale hydraulic responses. Compared with conventional pore morphology (PM) methods, the PRP approach reduced computational processing time by approximately 16–50%, while maintaining high predictive accuracy. Owing to its computational efficiency and robust representation of pore-scale characteristics, the proposed method provides a reliable basis for automated extraction of microstructural parameters. It offers significant potential for future cross-scale integration with discrete-element-method frameworks.