Predicting soil-water retention curve based on NMR technology and fractal theory

It has been recognized that the soil-water retention curve (SWRC) is a fundamental constitutive relationship of unsaturated soil mechanics, which plays a decisive role in the researches on the strength, volume change and permeability coefficient of unsaturated soils. The direct experimental methods for measuring the SWRC are time-consuming, and they usually provide some discrete data points which are short of sufficient continuity, thus they cannot completely meet the needs of researches on unsaturated soils. So, it is necessary to fast obtain the complete mathematical expression for SWRCs by using the indirect method. Based on the non-invasive nuclear magnetic resonance (NMR) technology, the relationship between matric suction ψ and relaxation time T₂ is established by using the Young-Laplace theory. Then, by using the fractal theory, an expression for the relationship between mass water content w and relaxation time T₂ is derived. On this basis, a specific NMR-fractal method is promoted to predict SWRCs. The experimental results show good agreement with the predicted values by the proposed fractal method.