Low field NMR based relative permeability and drying model for unsaturated granular materials

Abstract

Climate change and airflow variations critically influence soil-atmosphere interactions and subsurface evaporation processes. This study investigates the role of particle and pore sizes in the drying dynamics of unsaturated coarse-grained granular media under low-humidity airflow, employing coupled nuclear magnetic resonance NMR and matric suction measurements. Comprehensive experiments analyze grain size impacts on (1) drying rates, (2) matric suction evolution, and (3) T₂ relaxation times. Results reveal that finer-grained specimens retain higher asymptotic saturation due to stronger capillary forces, while smaller mean grain sizes d₅₀ correlate with elevated matric suction in constant-suction zones. A geometric constant (a = 4), derived from mercury intrusion porosimetry, BET specific surface analysis and NMR spectroscopy with various coarse-grained materials, enables direct conversion of T₂ relaxation times to pore diameter d_p and assuming simple cubic packing allow to estimate d₅₀. This constant underpins a log-normal pore size distribution model that aligns with suction measurements. We propose a three-stage drying model integrating surface evaporation, capillary flow, and soil-internal vapor diffusion, validated against experimental data for glass beads and sand. NMR-derived hydraulic properties enable accurate predictions of drying curves, advancing non-destructive characterization of unsaturated soils for geotechnical applications.