Abstract:

To address the issues of accurate conversion of the T₂ spectra at different temperatures and the definition of water retention states in low-field nuclear magnetic resonance (L-NMR) unfrozen water tests, this study conducted freezing L-NMR tests on saturated and centrifuged samples. The evolution characteristics of unfrozen water under two conditions were studied, achieving an accurate conversion of T₂ spectra during the freezing process and a precise delineation of the water distribution states. Building upon these insights, the migration mechanisms of unfrozen water in both states were subsequently analyzed. The main conclusions are as follows: (1) A temperature correction method for the freezing T₂ spectra was developed. Using this method, the disparities in the soil freezing characteristic curves (SFCCs) between the two states were scrutinized. It was observed that the unfrozen water content in the saturated samples was consistently higher than that in the centrifuged samples. The variance in water occurrence was pinpointed as the primary contributor to this discrepancy, and the T₂ threshold ranges corresponding to different water retention states were established as follows: film water (0.038−0.396 ms), capillary water (0.396−2.319 ms), and free water (≥2.319 ms); (2) Leveraging the evolutionary traits of unfrozen water and the Gibbs-Thomson effect, the surface relaxation rate p₂ = 49.49 nm/ms was ascertained as suitable for this experimental context. This facilitated the transformation from T₂ distribution to pore size distribution, and r = 58.8 nm and r = 344.0 nm were identified as the boundary pore sizes delineating film water, capillary water, and free water, respectively.   (3) The experiments revealed a temperature threshold at which the migration of film water ceases in both saturated and centrifuged samples. Additionally, the temperature threshold for saturated samples was found to be higher than that for centrifuged samples. It was concluded that the migration cutoff temperature for film water is controlled by the pore water saturation in the capillaries, with samples having higher saturation showing a higher migration cutoff temperature for film water. Both saturated and centrifuged samples demonstrated temperature thresholds at which the migration of film water ceased, with the threshold for saturated samples being elevated compared to that of centrifuged samples. The cessation temperature for film water migration is governed by the pore water filling rate within the capillaries, with samples exhibiting higher filling rates displaying elevated cessation temperatures for film water migration.

Key words: unfrozen water, water migration, low-field nuclear magnetic resonance (L-NMR), pore water distribution, soil freezing characteristic curve (SFCC)